The Eurypterida of New York/Volume 1/Morphology, anatomy, and terminology


THE EURYPTERIDA

I
MORPHOLOGY, ANATOMY AND TERMINOLOGY

General form. The body of the eurypterids, as a rule, is elongated, and often somewhat fishlike in dorsal view, but it may also become distinctly scorpioid. The slender fishlike form is typically expressed by Hughmilleria [see restoration, pl. 59] where the body expands but slightly in the anterior third and then tapers very gradually to the caudal spine; the head shield is semiovoid and the body less depressed than in other genera. Pterygotus is similarly built, but the head shield is more rounded in front and some species are broad and plump. In Eurypterus the lateral expansion or flattening of the body becomes manifest in both head shield and abdomen and the contraction of the body to the caudal portion is more abrupt. Stylonurus has a slender body which, however, expands gradually beyond the middle and then contracts more rapidly. In Slimonia a scorpioid appearance is produced by the squarish head shield and the long tubular caudal portion of the abdomen. An extreme differentiation from the slender terete body of Hughmilleria is reached in Eusarcus with its triangular head shield, broad flat body with subcircular outline, from which the long narrow tail is sharply set off.

We shall recur to this variant expression of the body and its bearing on the mode of life of these animals after a consideration of the appendages which are correlated to the form of the body and corroborate the evidence from the latter.

Integument. The body is covered by a chitinous exoskeleton which alone is preserved in the rocks and usually compressed into an extremely tenuous, carbonaceous, more or less wrinkled film. Notwithstanding its thinness it must have been, like the tough leathery integument of Limulus, very strong and able to furnish a stout basis for the powerful muscles of the creatures.

Figure 2 Diagram of an Eurypterid

a; Ia-VIa = appendages or legs o = "larval" median eyes or ocelli
Ab = abdomen om = ocellar mound
bl = balancing leg on = ocular node
ca = carapace or head shield op = operculum or first sternite
car = carina or telson pa = palette
Ce = cephalothorax or prosoma Po = postabdomen or metasoma
ch = cherlicerae, Ia, preoral appengdages or mandibles pp = pentagonal pieces
cj = coxa Pr = preabdomen or mesosoma
end = postoral appendages or endognathites ps, I-VI = caudal or postabdominal segments
ep = epimera or pleura of tergite r = rim
ga = genal angle rh = rhachis
gea = genital appendage (female) s, II-V = sternites
gl = glabella sf = swimming leg
gn = gnathobase t = telson
le = lateral or compound eyes t, I-VI = tergites
md = marginal doublure ta = tubular organs
me = metastoma or postoral plate tl = triangular lobes
mo = mouth va = visual area
mp = marginal plate on under side of cephalothorax wl 1-3 = walking legs

Probably the majority of the remains found are the cast exuviae from the frequent moltings of growing individuals. The recent Limulus may take as long as eight years to reach maturity when, with the last molt, the clasping organs of the males appear. It is therefore probable that the eurypterids also were relatively slow in growth and it is a fair presumption that the great majority of the specimens found represent immature individuals. They are mostly dismembered and among the fragments of the integument, head shields with the first body segment attached are especially frequent.

Figure 3 Eurypterus fischeri Eichwald
Diagrammatic longitudinal median section. a, carapace; b, metastoma; 1-6, tergites; I-V, sternites; 7-12, postabdominal segments. (From Schmidt)

We consider it possible, or even probable, that the molting took place as in Limulus through a rent formed back of the frontal doublure of the head shield, through which the animal crawled out. Not only have specimens been found, as the type of plate 6, figure 1, where there is a gaping rent along the front edge of the head shield, but it is also inconceivable that the animal could otherwise have freed its legs so as to pull itself out of its old integument.

The test was thicker on the head shield for that is always less wrinkled and has retained its form better, and it was on the whole, thicker on the dorsal than on the ventral side. The appendages were also clothed with thicker test and the basal segments which attended to the mastication of the food were furnished with an especially thick cover. The postoral plate was likewise thick and is always well preserved.

Scales. The test or exoskeleton is characteristically ornamented by scale markings. On the head shield these consist of tubercles which are mostly simple thickenings of the test, but in some forms, as Eurypterus pustulosus, may grow out into large hollow wartlike excrescences. On the abdomen the scales are sometimes disk-shaped but more frequently v-shaped or crescentic, rising in posterior direction.

Nieszkowski [1859, p. 335] has suggested that these scales are the attachment places of muscles, a view which to us seems to explain most peculiarities of their structure and distribution. The following points seem to support Nieszkowski's assertion:

1 The scales are more distinctly outlined on the inside than on the outside of the test. In some cases they are well seen on the interior while faint or invisible on the outside, as in the specimens of Eurypterus pittsfordensis.

2 On the abdominal segments they are distributed in distinct transverse zones or belts [see Eusarcus scorpionis, pl. 31; Pterygotus, pl. 79] which correspond to the muscle bands that function in moving the segments, and they are entirely absent over the anterior and posterior doublures where no muscles could be placed; on the dorsal side they are arranged in longitudinal rows that follow the course of the intestinal canal and indicate the insertion of muscles with a suspensory function.

3 They are especially strong and distinct on the movable plates of the underside of the abdomen which carried the gill plates and were strongly shifted forward and back in breathing and swimming.

In the mature Limulus the crust is too thick to show any such muscle impressions on the outside, and moreover, the body has become separated into three solid fused regions (the cephalothorax, abdomen and telson) which only are movable upon each other. The muscles have thus become localized and fastened to strong internal processes or entapophyses. No such entapophyses have been found in the integument of the eurypterid abdomen, and there is in these structures evidence of the primitive condition of the musculature indicating a state of dissolution into many small muscular fasciculi. The thin-shelled young Limulus, less removed from the distinct abdominal segmentation of the embryonic stage, still exhibits a like distribution of the muscles and also shows from the outside such attachment scars as the eurypterids. Even in the mature Limulus, the extensors, or abductors, and the flexors, or adductors, of the gillbearing ventral plates are still subdivided into many radiating fasciculi, and likewise the extensors and flexors of the tail spine arise from a large number of fasciculi.[1]

The longitudinal zones or series of scales are especially distinct in Eurypterus remipes [pl. 8, fig. 5] and lacustris. They inclose here a distinct smooth, flat median zone, which corresponds to the course of the intestine and there is reason to believe that this double series of scales marks the bases of the muscles and suspensory organs of the intestine. In Pterygotus the distal portion of the intestine was sharply bent and the angulation has produced a distinct protrusion on the dorsal side of the last segment [pl. 75].

4 The scales are not found on the telson spine, except the anterior swollen portion where the extensors and flexors were attached, nor on the broad, solid oarplates of the swimming legs, while they are present on the leg segments, frequently in distinct longitudinal rows which in some cases develop into regular continuous entapophyses on the inside of the integument, as in the powerful arms of Dolichopterus [pl. 43].

5 On the postoral lip (metastoma) the scars are most strongly developed on the side that is turned toward the body and held by muscles.

6 The prevailing form of the scales is that of a V or a crescent pointing backward, and this shape, together with the fact that the integument is distinctly thicker and darker along the edges of the scales, is indicative of their function, since the muscle scars here, as among mollusk shells, are marked by their thickened edges. Forms with thicker integument, as Eurypterus microphthalmus, do not exhibit any scaly sculpturing on the outside.

7 The scales are well differentiated from the tubercles which are scattered among them [see test of Eusarcus, pl. 35, fig. 4] and on those parts where the thickness of the integument prevents the appearance of scales over the surface, as on the carapace, these tubercles alone are observed. Similarly the serrations of the edges of the rudder plate of the swimming legs, of the epimera of the abdominal segments and of the telson are independent of scales. But it appears that the spinules on the dorsal crust originate from scales, we surmise from a lengthening of the point of the angle of the V. In Eurypterus remipes, lacustris and pittsfordensis, as well as fischeri, but a single transverse row of such spines is developed on each segment. In E. remipes the longitudinal series of scales on the postabdomen over the intestine show a strong tendency to develop into overlapping spinules and in the Carbonic subgenus Anthraconectes all scales possess this tendency.

Besides the scales and spines the integument of the eurypterids was also in many parts covered with fine hairs as in Limulus, especially on the lips and legs. These have been observed and photographed by Holm in E. fischeri.

Pores of doublures. Schmidt [loc. cit. p. 67] has observed that in Pterygotus the posterior third of the dorsal thoracic segments is provided with a separate interior lamella or doublure ("Umschlag") perforated by numerous fine tubes especially abundant toward the posterior and lateral margins. In Eurypterus this doublure is present but imperforate.

A slightly weathered specimen of Pterygotus buffaloensis very beautifully exhibits this interesting feature not only verifying Schmidt's observations but also allowing some amplification of them. In this specimen, the pores are distributed over the whole lamella, but are most closely arranged and most distinct in a belt occupying the posterior third of the lamella except for the hindmost millimeter which is entirely imperforate. Where fully exposed, they appear as sharp black lines, about 11/2 millimeters long, passing from below and posteriorly, obliquely upward and forward. At the posterior margin of the belt they stand more vertical but they rapidly bend forward and increase in obliquity of direction toward the anterior margin of the belt. They are most distinct on the tergites, but have been seen on the sternites in a similarly situated, though narrower belt, corresponding to the smaller width of the doublure. Similar belts of pores are visible on the posterior thirds of the broad doublures of both the dorsal and ventral sides of the abdominal segments, and on the narrow doublure of the posterior margin of the head.

The weathered head of the same specimen also exhibits anastomosing lines near the center and a ring of several rows of black points and lines which seem to indicate either the presence of similar pores or of a system of pits.

Another row of pits which seem to have been pores, is also seen just inside the flexure of the head shield into the doublure. These would seem to correspond to the row of "Punkte" observed by Holm in Eurypterus fischeri. Such minute pores or canals may have given passage to sense organs (sensory setae or bristles) or to tegumentary glands.

Serrations and spines. The edges of the integumental plates are frequently furnished with sharp serrations, like those of the abdomen of Limulus. These serrations are especially frequent on the anterior edge of the palette of the swimming leg, on the epimera of the postabdominal segments and on the lateral edges of the tail spine. In Eurypterus the distal edges of the segments of the swimming legs are also distinctly serrate. Frequently an alternating series of larger and smaller serrae is found. A good example of this is the telson of Pterygotus buffaloensis.

The spines on the legs and those on the manducatory edges of the coxal segments have originated from fine hairs, which grew first into strong bristles and finally into spines. These spines become powerful spurs in some genera, as Eusarcus and Stylonurus. In Ctenopterus, the subgenus of Stylonurus, they are sometimes broadened and may have assisted in swimming. Like the hairs, the spines are inserted and fixed by ringlike swellings of the integument at their bases and frequently broken out, leaving distinct scars.

Cephalothorax. The body of the Eurypterida consists of 18 segments and is functionally divided into two sharply distinct regions, the cephalothorax or prosoma, consisting of six fused segments and the abdomen consisting of 12 segments. The cephalothorax with its appendages is specialized for locomotion and prehension, while the abdomen is essentially a vegetative sac. The six segments of the cephalothorax find their expression in the six pairs of appendages. The cephalothorax consists of the carapace or head shield and the organs of the underside.

The carapace is of small size, one fifth of the body (exclusive of the telson) or less in length. This small size is a feature that gives the eurypterids a greatly different aspect from Limulus with its immense semicircular carapace, but it well corresponds to the relative dimensions in the scorpions. In Limulus it is expanded so much both laterally and frontally that it entirely covers the legs and is the result of extreme adaptation to a mud-groveling habit, but in the eurypterids it never covers more than the basal segments of most legs. It nevertheless shows considerable variation in relative size, which is roughly correlated to the size and weight of the legs. Thus Dolichopterus with its very stout walking legs and extremely long swimming legs has a relatively very large carapace, while Pterygotus with thin walking and short swimming legs has, in spite of the enormously extended chelicerae, a remarkably small carapace.

The carapace is typically subrectangular in Eurypterus, Dolichopterus, Slimonia and Stylonurus, semiovoid to semicircular in Strabops and Hughmilleria, and subtriangular in Eusarcus.[2] The broadly semielliptic or semicircular form is manifestly original and primitive, as indicated both by the larval stages of the eurypterids and the carapace of the Cambric Strabops, while the subrectangular and subtriangular forms are the extremes of different lines of development.

The carapace culminates at the middle or the posterior third in the median ocellar tubercle. Along the margin it is more or less flattened and the border is frequently thickened and beveled, forming a shoveling edge. This is notably the case in Eurypterus but not in the Pterygotus branch of the subclass.

In the great majority of specimens the surface of the carapace is flattened by compression. Nieszkowski figured in Eurypterus a short ridge extending backward from the middle of the frontal margin and two crescentlike lateral ridges on which the compound eyes are situated. Schmidt says that Nieszkowski exaggerated these ridges. He himself describes a narrow pointedly triangular prominence reaching from the posterior margin forward to the ocellar tubercle, whence two broad sector-like elevations extend forward and inclose a median depression extending to the frontal margin. Holm has considered the ridges as accidental and represented the carapace of Eurypterus as evenly rounded with the exception of the frontal elevated areas.

The usual waterlime material does not give any evidence of the original form of the carapace of Eurypterus, the integument having been completely flattened out in the fine mud. A sandy dolomite bed of the Bertie waterlime at Morganville, Genesee co., has however furnished a few specimens of Eurypterus remipes that are uncompressed. These [pl. 6, fig. 6] have glabellalike posterior median ridges well defined by two subparallel furrows deepest half way between the posterior

Figure 5 Transverse section through the head of Limulus
ht, heart; liv, liver; es, entosternon; end, endognathite. (From Packard)

margin and the ocellar tubercle, which, by the way, the glabella does not quite reach. They also show a crest extending from the lateral eyes to the posterior margin and separating the elevated apical area from the steeply outward sloping lateral areas. This sculpturing is so remarkably like that of Limulus that we have no doubt it represents a general feature of the eurypterids. The dorsal furrows bounding the glabella correspond to entapophyses, serving for the attachment of muscles. The appended sections of Limulus [text fig. 5, 6] show the relation of the glabella to the position and extension of the heart and the relation of the glabellar furrows to the muscles holding in place an internal cartilaginous plate, termed the endocranium, plastron or entostemon,[3] which in its turn forms the fulcrum for other important muscles. Owen has shown that the glabellar furrows, or rather the entapophyses or infoldings to which the furrows correspond, are the bases of the powerful levators of the preabdomen and also of the muscles which serve to steady the entosternon while the latter furnishes the fixed points for the flexors or depressors of the preabdomen and the important muscles that hold and move the coxal joints of the legs. [graphic]

Figure 6 Longitudinal section of Limulus
liv, liver; pr, proventriculus; st, stomach; hr, heart; cp, cartilaginous plate, entosternon; int, intestine; a, anus; br, brain. (From Packard)

The glabellar furrows of Eurypterus thus correspond quite precisely in position and extension to those of Limulus and the appendages on the ventral side are entirely homologous and of like position and structure; hence it is to be inferred that the same system of muscles existed in Eurypterus as in Limulus, that also Eurypterus may have possessed a cartilaginous entosternon and that the glabellar furrows served as bases for the levators of the preabdomen and the lateral levators of the entosternon. These furrows are also well developed in species of Stylonurus, as S. cestrotus.

This glabellar ridge with its bounding furrows is most distinctly marked in the embryonic Limulus [text fig. 24] where it bears the ocelli at its anterior point. It there continues over the abdomen and gives this stage its well known trilobitic appearance. It is still more prominent in the fossil Limuli (e. g. L. walchi), than in the recent species and is therefore apparently an old character possibly inherited from the common ancestor of the merostomes.

The elevated area of the glabella manifestly served, as in Limulus, to receive the anterior portion of the heart, while the space between the carapace and the ventral membrane in front of the ocelli contained the liver. We find that the interior surface of the carapace sometimes exhibits in this region, in E. lacustris and remipes, anastomosing radiating lines similar to those seen in many trilobites (Conocephalites, Harpides, etc.), and which have been interpreted by Jaekel as liver impressions.

The edge of the carapace is bent under, forming a doublure. This is usually narrow as in Eurypterus and runs out toward the genal angles. In Stylonurus, however, and notably S. myops [pl. 52, fig. 10] it becomes very broad and concentrically striated. In Limulus the doublure broadens in the median part of the anterior portion into a concave triangular shield that is said to serve as an inlet to the water for respiration when the broad carapace is resting on the mud. A triangular area of like relation and relative dimensions is set off in some species of Stylonurus [pl. 46, fig. 11] and may have had a like function.

To the doublure are attached by an open suture two plates, figured by Hall [1859. v. 3, pl. 80A, fig. 12], meeting in the median line along a suture and together forming a horseshoe-shaped organ, which toward the mouth passes gradually and by an irregular contact into the very thin membrane that surrounds the coxal segments of the legs. One of these marginal plates is shown in place in plate 5, figure 6, where the carapace is partly removed. Frequently in specimens of Eurypterus remipes not completely flattened, they have prevented the marginal portion of the carapace from further sinking in and therefore appear on the compressed carapace as a very marked smooth, fiat marginal zone [pl. 6, fig. 5]. In Pterygotus a third plate is intercalated between the two marginal plates in front of the mouth, forming an epistoma that occupies the same position as the hypostoma of the trilobites.

Limulus possesses in front of the chelicerae a wartlike node on the ventral membrane which has been shown by Patten [1894] to contain an olfactory organ. While Holm's photographs demonstrate the absence of anything similar in Eurypterus fischeri, it seems to us that nodes observed in specimens of Eusarcus and Stylonurus in the corresponding place may possibly indicate the presence of a like organ in these genera.

The connection of the carapace with the abdomen is accomplished by articulations near the postlateral angles, well seen in plate 6, figure 5. It is indicated by an abrupt change in the direction of the posterior margin where the truncation of the genal angle begins. Between the articulations the margin of the carapace is curved forward, so that an open slitlike space remains between the carapace and the abdomen which is occupied by a thin membrane connecting the doublures of the carapace and first abdominal segment. As Holm has pointed out, the open space indicates that the movability of the articulation between the carapace and the abdomen must have been very considerable.

Eyes. The carapace bears two pairs of eyes, the large lateral or compound eyes and the median eyes or ocelli. The lateral eyes distinctly fall into two groups by virtue of structure and position.

The first of these groups exhibits a smooth visual area which is more or less crescent-shaped and borne on an elevated ocular node between the glabella and the lateral margins. This type of eye is exemplified by Eurypterus and is also found in Dolichopterus, Drepanopterus and Stylonurus. Probably Strabops also possessed eyes of this type; Eusarcus, in regard to its eyes as well as its whole body form, is an aberrant type, for while it possesses apparently smooth bean-shaped eyes these are marginal as in those of the second group, being borne near the anterior angle of the subtriangular carapace.

The second group of eyes is typically represented by Pterygotus, and found also in the genera Slimonia and Hughmilleria. In these the visual area occupies the whole node, is marginal and faceted. In Pterygotus the immense eyes [pl. 73] occupy the antelateral angles. The side of the somewhat globose carapace is impendent in the antelateral region so that in flattened individuals part of the lateral eyes is pressed over on the underside (particularly well seen in P. bilobus); similarly in Slimonia, where the lateral eyes lie at the anterior angles of the rectangular carapace, half of the eye is on the underside [see Woodward's restoration, 1872, pl. 20]. The connection between the Pterygotus and Eurypterus eyes is afforded by that of Hughmilleria as we show in the generic discussion of that genus. While it is marginal in H. socialis, the genotype, it is still submarginal in H. shawangunk and while it is smooth on the outside, it shows delicate facettae on the inside.

Homology of lateral eye in Pterygotus and Limulus. No observers have recorded the presence of corneal facets in the eyes of Eurypterus and Holm states that even in excellent microscopic preparations he has been unable to notice anything but an apparently smooth, uniformly thick cornea; the faceted eyes of Pterygotus, however, have been known to the earliest writers on this group of fossils. Beyond the fact that the lateral eyes of Pterygotus possess faceted corneae, nothing can be gleaned from the literature, and the figures given are equally inconclusive showing either projecting round lenses in a square meshed interstitial test or sclera [cf. Huxley & Salter Monogr. I, pl. 3, fig. 1b] or lenticular depressions with a hexagonal scleral test [Woodward, pt. 2, p. 56]. More frequently it is stated that the facets are not discernible, a fact attributed to their extremely small size.

Several carapaces of Pterygotus macrophthalmus and P. buffaloensis, their visual surfaces excellently retained, permit us to elaborate the anatomy of this type of eye with a fair degree of precision, and to demonstrate the entire homology in structure of the eyes of Pterygotus and Limulus. The visual surfaces of the specimens in question exhibit five distinct states of preservation, here illustrated by diagrams.


Figure 7 Compound eye of Pterygotus; diagrams of preservation states
In states I and II the visual surface is perfectly smooth without a trace of reticulation or lenticular depressions or prominences even under condensed light and under water. Two specimens show this state in the exterior view, two more as intaglios of the exterior of the head. Among the former is one of the best preserved carapaces [pl. 69, fig. 7] which distinctly shows the fine granulation of the surface and hence might be expected to retain also the finest details of the visual surface. The other specimens with smooth eyes are so well preserved that the corneal facets should be visible. Where the smooth surface is seen in an exterior view, the presence of the carbonaceous film indicates that the exterior of the test is actually under observation while in case II where the smooth visual surface is seen in a cast, the absence of the test proves likewise that we do not have before us an interior view of that surface.

The Second group comprises preservation states III and IV. In Case III the visual surface in an exterior view exhibits lenticular depressions in a network of squarish meshes. This is well shown in the large head [pl. 73, fig. 1]. In case IV the visual surface is seen in a cast and provided with a system of low papillae corresponding to the depressions seen in case III. Here the test is reduced to a mere carbonaceous film.

The relation between these two states of preservation, the absolutely smooth and the papillate, is indicated in the condition V exhibited by the fine head shield in the Buffalo collection. Here is a smooth visual surface showing whitish circles in a dark reticulate mass and resting on this surface patches of a smooth carbonaceous test which in this case shows but the faintest shadow of the underlying lenticular structure.

On plate 72, figure 1, we illustrate an internal mold of P. buffaloensis from the collections of the National Museum and the external mold of the same specimen from the Museum of the Buffalo Society of Natural Sciences is shown on plate 72, figure 2. In this external mold the test of the lateral eye is radially wrinkled and lacks all traces of facets, while in the internal cast the eyes are finely faceted but without radiating wrinkles. In this case, there must have taken place before the entombment of the specimen a partial separation and a wrinkling of an outer smooth cornea.

We may hence conclude with entire safety that this Pterygotus at least possessed a smooth, relatively and uniformly thick cornea and below this a system of lenses.

In accordance with this conclusion the lenses were either separate from the overlying cornea or they were only papillary prolongations of its underside. If the former they have a structure like that of the holochroal eye of the trilobites; if the latter then the structure is in entire accordance with that of the eye of Limulus [see text figures 8 and 9].

The choice of the alternatives seems to be indicated by states IV and V; for both can only be explained by assuming that the lenses have been lifted out as a whole or system, leaving the sclera forming the sockets. In case III the epidermal layer into which the lenses projected is still preserved; in case IV this is lost and only the impression left. In the latter case the cornea and the attached lenses must have been lost before burial by the sediment as otherwise the papillate cast should not have been produced. It is obvious that if the lenses had been distinct and separate from the cornea, we would find, after the loss of the cornea, the majority of the lenses still embedded in the epidermal layer, while in fact in these two cases none at all are thus preserved.

The sole case which might be taken to exhibit separate lenses is the last (V), where a smooth faceted surface is seen. As stated before, the lenses appear as light circular spots in the brown carbonaceous test. Close examination shows that they consist of semilenticular dolomite fillings of depressions that correspond with those observed in case III. They could be taken either as demonstrating that the cornea with the papillae was lifted out of the sclera before the burial of the specimen in the sediment and thereafter the depressions filled with mud, except where patches of the cornea adhered to the eye; or as representing the filling of corneal cavities which function as lenses and are homologous to the anterior corneal cavity observed by Clarke [1888, p. 258] in the schizochroal eye of Phacops rana. It is obvious that the former explanation is, in view of the preservation states III and IV, by far the more plausible, especially since the presence of a continuous smooth cornea precludes the comparison with the schizochroal trilobite eye.[4] There are besides these direct arguments for the limuloid structure of the Pterygotus eye, other facts which point, though less directly, to the same conclusion. These are the distinct continuity and at the same time strange tenuity of the large cornea of Pterygotus in contrast with the solid holochroal eye of such a trilobite as Asaphus. These characters show themselves in its wrinkling (often distinctly radiate, as in plate 72, fig. 2, more often concentric) or bursting in other specimens. That this cornea is continuous with the integument of the head is indicated by the fact that it is not divided from it by any distinct line and in macerated heads does not separate from it.

This evidence may be supplemented by two a priori reasons for similarity of eye structure in the eurypterids and Limulus: a) in all other organs these organisms have been found to agree with that ancient genus, b) the eyes of Limulus are of a remarkably primitive type such as would actually be expected in these archaic arthropods.

In view of this inferred structural identity in the lateral eyes of Pterygotus and Limulus, the stage of development which the latter has attained becomes a matter of interest to us here.

Packard [Am. Nat. 1880, 14:212] describes the eye of Limulus as follows:

The structure of the eye is very unlike that of any other arthropod eye. The cornea is simply a smooth convex portion of the integument, which is much thinner than the adjoining part of the chitinous skin. There are no facets, the cornea externally being structureless, simply laminated like the rest of the integument. In the internal side of the cornea are a series of solid chitinous conical bodies, separated from one another by a slight interspace and in form resembling so many minié-rifle balls; the conical ends of these solid cones project free into the interior of the body, and are enveloped in a dense layer of black pigment. Within the base of these cones are secondary shallow cuplike bodies or shallow secondary cones. It is these primary cones which, seen through the smooth convex translucent cornea, give the appearance of a faceted surface to the external eye.

All the parts thus far described except the pigment layer, are molted with the rest of the crust, and the large slender cones can be easily seen by viewing a piece of the cast-off eye; the solid cones being seen projecting from the inner surface of the cast-off cornea.

The author adds: "So far as we can ascertain, no arthropod eye is so simple as that of Limulus."

Watase [Biol. Studies, Johns Hopkins Univ. 1880, 4:287] in consequence of his investigation of the compound eyes of arthropods considers the ommatidium of the lateral eye of Limulus as making the nearest approach to the primitive condition. "It is nothing more and nothing less," he states, "than a depression in the skin, with the thickened chitinous cuticle fitting in the open cavity and acting as a lens to condense the light." We have copied two of Watase's excellent figures to illustrate this structure [text figs. 8, 9].

In view of the close relationship clearly uniting Pterygotus with the other eurypterids, notably Eurypterus and Eusarcus, we have little reason to doubt that the lateral eyes in all were of like structure. Nevertheless, there may here exist a difference from that of Pterygotus, for in the other eurypterids no facets have been observed. It is obvious that both Eurypterus and Eusarcus possessed an exteriorly smooth cornea just as Pterygotus; as to the interior of their compound eye, however, several possibilities present themselves: either the lenses were separate from the cornea and thereby lost in fossilization, not being united by the sclera as in the holochroal eyes of the trilobites; or the facets were so feebly

Figure 8 Limulus. Two ommatidia shown side by side, partly schematic. The thick unshaded body is the chitinous covering of the eye. L. lens cone, fitting into the depression of the skin. Rt. retinula. G. ganglion cell. (From Watase) Figure 9 Diagram of the compound eye of Limulus, the black, heavy line representing the ectoderm and each depression in this layer corresponding to an ommatidium. (From Watase)

developed as to escape observation. There are no records as far as we are aware, either among living forms, or among the fossil merostomes, to support the former hypothesis. Numerous recent crustaceans, among them the venerable Apus, either lack the facets entirely or have them so poorly developed that they are hardly noticeable,[5] and this fact seems well suited to shed light on the failure to detect the facets in Eurypterus and Eusarcus.

In some arthropods the crystalline cone assumes a transparent semiliquid state [see Watase, 1890, p. 147], and it may perhaps be assumed that the eye of Eurypterus had advanced one step beyond that of Pterygotus leading to the separation of the lens cone into an independent crystalline cone, and that this crystalline cone had failed to become hardened. In other arthropods, as in such insects as form Grenacher's "aconous type" of the compound eye, the whole ommatidial cell may remain as a clear transparent body [Watase, ibid].

The bearing of the position of the compound and simple eyes on the habits of life of the eurypterids will be noted in another chapter.

The median eyes or ocelli, which are also frequently termed the larval or simple eyes, consist of two single, transparent spots of the integument in round pits with ringlike walls situated on a small mound or tubercle. In Eurypterus and Eusarcus the transparent spots are distinctly thickened into lenslike bodies. The ocellar tubercle is always situated on the median line of the carapace and it wanders forward and backward with the lateral eyes so that as a rule it is on a cross line connecting the posterior extremities of the lateral eyes. Exceptions are made by the genera Eusarcus, Pterygotus and Slimonia where the lateral eyes have become marginal at the front of the carapace while the ocelli remain on the middle of the carapace, thus retaining the advantage of their position at the apex of the shield. In Limulus, on the contrary, the lateral eyes have remained in the posterior half and the ocelli have wandered to the front.

As to the function of the ocelli and compound eyes, observations on other classes of arthropods which possess both groups have led to the conclusion that the two sets are complementary to each other, the compound eyes being adapted to distant sight while the ocelli are myopic, as indicated by their highly convex lenses. This specialization has been necessitated by the slight adaptability of the stiff compound lenses to different distances. The most primitive ocelli lack lenses entirely and are mere pigment spots sensitive only to intensity of light.

Appendages of cephalothorax. The cephalothorax or prosoma bears six pairs of limbs which are homologous with those of Limulus. It is this close homology and the detailed comparison which is possible between the limbs of Limulus and the eurypterids that are among the strongest proofs of their intimate relationship.


Figure 10 Limulus polyphemus; female, from ventral surface
ab, abdomen; an, anus; ch, chelicera; chi, chilarium; cp, cephalothorax; ol, olfactory organ; op, operculum; sp, spine. (From McMurrich)
The six pairs of limbs are currently divided into the preoral (the first) and postoral (the following five pairs). The preoral limbs are the chelicerae or mandibles, the postoral the walking, and burrowing or swimming legs. Besides these the mouth is surrounded by platelike appendages, functioning as lips. Those are the epistoma, endostoma and metastoma.

a The chelicerae in Limulus are small [see text fig. 10] and consist of three segments, the much compressed but relatively long basal segment and the two forming the pincers or chelae. The basal joint articulates, as Holm has pointed out, with an unpaired, lanceolate plate placed between the coxal segments of the first pair of walking legs.

The preoral appendages of the eurypterids exhibit great differentiation. They are smallest in Eurypterus, where they have been described in detail by Holm and have proved to have almost the exact structure of those in Limulus. We figure here the chelicerae of E. remipes and E. lacustris [plate 7, figure 1, and plate 12, figure 1] which verify Holm's observations. Those of Stylonurus have a like structure and similar relative size as shown by Hall and Clarke [see under S. excelsior]. Eusarcus had similar chelicerae which, however, were relatively large and had stronger chelae. The specimen of E. scorpionis reproduced on plate 32, retains them in their natural position. Woodward figured them, as pointed out by Laurie, in Eurypterus scorpioides, which is an Eusarcus. In Dolichopterus and Drepanopterus they are not known. In general it may, however, be said that they were extremely similar in the genera of the Eurypterus group, viz, Eurypterus, Eusarcus and Stylonurus.

On the other hand they show extreme variation in the genera of the Pterygotus group, viz, Hughmilleria, Slimonia and Pterygotus. In Slimonia they were discovered by Laurie [1893, p. 511, pl. 1, fig. 3] and found to be very small and with strong curved pincers, but otherwise as in Limulus. Those of Hughmilleria have been fully described by Sarle. They are larger [pl. 61, fig. 6] than in any other genus except Pterygotus, projecting beyond the margin of the carapace when extended and thus forming a transition to their condition in the latter genus. In Pterygotus they have been developed into the gigantic pincers which give that genus its fantastic aspect. Notwithstanding their great size in Pterygotus, there still prevails much doubt as to their structure. The current restoration [Zittel-Eastman's Textbook, fig. 1423] is one that would indicate a structure greatly different from that of the chelicerae of the other eurypterids. Our material fortunately sheds light on this problem and we have for this reason inserted here the following note.

The morphology and formation of the chelicerae in Pterygotus. In several specimens of P. macrophthalmus and P. buffaloensis at our disposal, the chelicerae are so excellently preserved as to remove all doubt regarding their morphology and function. In one [pl. 74, fig. 1] the chelicera of a large individual is perfectly preserved from the distal extremity of the chelae or pincers to the base of attachment. This shows distinctly that the organ consisted only of one unjointed long arm carrying the terminal pincers. In the second specimen the arms of both chelicerae lie side by side, while the pincers of both have swung back, one until it is subparallel with the arm. These latter chelicerae serve to suggest the ready reversibility of the pincers. On comparing the length of the pincers with that of the arm [specimen pl. 77, fig. 3] one finds that the pincers are as long as the uncontracted part of the arm. The contracted basal portion, which amounts to about one fifth of the length of the arm is, as our material and the drawings of the British specimens indicate, nearly always missing. In another specimen, this part contrasts by its thinness with the thicker test of the arm, which terminates abruptly at a convex line along the contracted part. One might at first glance infer the presence of an articulation at this point, but the continuance of the test and the contrast in thickness of it on the arm and of the basal contracted part show that this latter was rather of the nature of a membrane and probably a part of the epistoma, Its basal edge is ragged and obviously torn.

The corresponding lengths of the pincers and arms and the actual occurrence of pincers thrown back, demonstrate, we believe, the functional possibility and competency of the pincers to grasp food and carry it to the mouth. The crustaceans afford several instructive examples of analogous prehensile organs. One of these is the giant spider crab from Japan (Macrochirus kämpferi) seen now in many of the larger museums. It possesses a pair of immensely long prehensile chelate limbs which consist of two long segments and the chelae, besides a small proximal segment. In the articulation it can be doubled in the exact middle between the two equally long segments, thus serving to bring the prey readily within reach of the masticating edges of the other limbs surrounding the mouth.

In the restorations of Pterygotus, a varying number of segments has been assigned to the chelicerae. Salter and Huxley [op. cit. pl. 15, fig. 6] gave four segments (the supposed additional basal ones not visible in the dorsal view), but Woodward doubled this number in his figure [op. cit. pl. 8, fig. 1]; and the latter is retained by Schmidt [op. cit. p. 73, fig. 1B] and has long since entered the textbooks. While Salter and Huxley figure the chelicerae as rigidly straight, as they are indeed seen in our specimens, Woodward gave them a graceful backward curvature. Schmidt again drew them nearly straight.

A perusal of the literature suggests that this conception of the manyjointed composition of the chelicerae is based on doubtful evidence. Woodward, for example, says of P. anglicus [p. 37] that "three joints at least may be observed " and of P. bilobus [p. 57] "there appear to be five joints in the antennae of this species, but it is seldom that their true line of articulation can be readily distinguished. Five are clearly to be seen in one of the antennae figured in the accompanying woodcut, figure 10." Schmidt states [p. 74] "Von den Scheerenfühlern sind in meinem Material nur die letzten beiden eigentlichen Scheerenglieder vorhanden, die ersten Glieder, von denen 3–5 angenommen werden, fehlen uns bisher." Laurie [p. 516] concluded that the question of the number of segments is still unsettled, and that the markings resembling articulations on the proximal portion may be due to crumplings of the undoubtedly thin cuticle, adding: "I believe them to have consisted of three segments—a long straight proximal one, and the two distal ones which possess the toothed pincers."

Laurie's contention is corroborated by our examples of P. macrophthalmus and buffaloensis and its correctness is suggested by the rigidly straight direction of the chelicerae in both the European and American species which can only mean the lack of articulation in the long arms. Finally we have also been able to convince ourselves of the presence of but three articulations in a fine specimen of one of the principal English species, P. bilobus, in the possession of the American Museum of Natural History. In this specimen both chelicerae show long straight well preserved basal segments without any trace of articulation in them, the whole chelicera being distinctly composed of but three segments.

Another problem in regard to the chelicerae of Pterygotus not yet solved by direct observation is their point and mode of attachment to the cephalothorax. A glance at the restorations of P. anglicus by Woodward, and P. osiliensis by Schmidt; both reproduced in Zittel's Handbook and in textbooks, will make this point clear. In the former they are represented as attached to the foremost point of the underside of the cephalothorax; in the latter they appear as inserted just in front of the mouth at the base of the epistoma. Schmidt says regarding this point [p. 73]: "Um auf die Scheerenfühler zurück zu kommen, so hätte ich ihren Ansatzpunkt gern, wie meine Vorgänger, nach vorn an der Unterseite des Kopfes verlegt, aber die eben besprochene 3-theilige Umschlagsplatte liess eine andre Auffassung nicht zu, als ich sie oben auseinandergesetzt. An ihrer Oberflache ist nirgend ein Platz für den Ansatz der Scheerenfühler und zugleich war für diese Umschlagsplatte selbst keine andre Deutung möglich." Laurie has not critically discussed this question in his study of the eurypterids, but only suggests that there was perhaps no properly developed articulation between them and the epistoma because they are always found torn off.

While we are not in a position to offer direct evidence on this problem, or at least only such as is inconclusive, we believe in the correctness of Schmidt's inference for the following reasons. One of our specimens shows that the six appendages of one side radiate from one point, which would mean that the chelicerae were inserted directly in front of the first pair of walking legs. But aside from this observation, the fact that the attachment of the chelicerae, in Slimonia, Eurypterus and Hughmilleria, has since been fully established to be directly in front of the mouth as in Limulus, leaves little doubt that the large pincers of Pterygotus, if they are at all homologous to the minute chelicerae of those three genera, must have had the same place of insertion, viz, at the posterior end of the epistoma. We may add that the homology of the large pincers of Pterygotus with the chelicerae of Limulus, Eurypterus etc. would also suggest as an a priori conclusion, their composition of but three segments and the lack of a distinct articulation with the epistoma.

The general conclusion from the foregoing observations is that the chelicerae exhibit a remarkable identity of structure in all genera despite their great differences in relative size.

b Postoral appendages. Of the five pairs of postoral limbs which are frequently designated as endognaths or endognathites, the first and fifth show the greatest amount of differentiation, while the intervening three pairs, as a rule, are very much alike and are functionally uniform, mostly serving as walking legs. We find the same condition in both Limulus and the scorpions and may therefore infer that the intermediate legs retain the original condition and that differentiation most easily affected the most exposed pairs, the first and last.

Generally all legs increase in length regularly from in front backward. This condition is typically shown in the more primitive genera Drepanopterus, Eurypterus and Hughmilleria.

The legs of Drepanopterus [pl. 54] exhibit the least differentiation of all genera, whose legs are known. All five pairs form a series of limbs which increase in length backward and are very much alike. One distinction, however, is that the first three are provided with a pair of fairly large spines on each segment, the posterior being longer than the anterior, an arrangement which was obviously of great assistance in pushing the body forward. The spines become gradually reduced backward in the series of legs; in the third pair posterior spines only are still well seen, the anterior ones being reduced to mucros. On the last pair they are all reduced to mucros. From the legs of Drepanopterus those of Stylonurus can be directly derived.

In Hughmilleria the first four pairs form continuous series of walking legs, all four being equally spiniferous and undifferentiated.

In Eurypterus the structure of the legs has been most minutely described by Schmidt and Holm in E. fischeri, and our large collection of E. remipes and lacustris corroborates their excellent work. The first three pairs are thick and heavy and increase in length regularly backward. The first leg consists of seven, the second and third of eight segments each, the terminal claw included. They are convex on the upper and flat on the underside and so articulated that they can be bent only downward. The principal spines are articulated and paired, the posterior one of each pair much longer than the anterior, an arrangement that also aided in pushing the body forward. The fourth pair shows a different form; it is comparatively slender, its segments are flat and lack the long spines, except the penultimate segment whose two spines, together with the terminal spine (forming a ninth segment), make a flat extension of the leg in the plane of the greatest length. Holm pointed out that this leg had its principal articulation between the basal and second segments. This fact and the form indicate that it aided in swimming, but probably had as its principal function the balancing of the animal in swimming.[6] This differentiation of the fourth pair of legs for another function than that of walking is most distinct in the genus Eurypterus.

Dolichopterus [pl. 40] is most nearly like Eurypterus in the character of the first four pairs of postoral appendages. They form a similar series, with the difference, however, that the fourth pair is considerably longer than the third. The first to third pairs are stouter than in Eurypterus, the spines much longer and the spines of each pair of subequal size. The fourth pair is still better adapted to its swimming and balancing function through the greater length of the leg, the greater breadth of the segments, and especially the lobelike character of the spines of the eighth segment which clearly exhibit the tendency of the leg to enlarge its lateral surface.

Stylonurus represents the extreme end of a branch that has developed through Drepanopterus. Its legs are hence to be regarded as derived from those of that genus. As restored by Woodward and by Beecher [see under Stylonurus] the first three pairs were conceived as short and spiniferous, while the fourth and fifth pairs were enormously extended, of subequal length and without spines. From the observations of the writers on the species from Otisville (S. cestrotus) which exhibits the last four pairs of legs, these formed a continuously increasing series of long legs, giving the specimen very much the aspect of a spider crab. The first two of these were, however, furnished with a great number of paired spines or leaflike appendages of the underside. It is therefore concluded that in this group, for which the term Ctenopterus is here proposed, a distinct differentiation of the legs has taken place into three spiniferous anterior and two nonspiniferous posterior pairs. Laurie's Stylonurus elegans is a form in which a like series of at least four pairs of long legs is shown, the two foremost of which bear many long spines [see text fig. 62]. In Stylonurus proper, as represented by S. logani, the first three pairs of legs appear to have retained more of their original character, in being relatively shorter and bearing only one pair of spines on each segment. Still another type of differentiation, not represented in our rocks, is shown in S. scoticus Woodward.

Eusarcus [pl. 27] represents a distinctly aberrant line of leg development corresponding to the entirely peculiar structure of the animal. The first pair of legs is of the length and character of that of Eurypterus; the second to fourth, however, form a series that decreases in size backward, the second being the longest of the walking legs. Correlated with this marked difference from Eurypterus is the greater length of the anterior spines on each leg. It is manifest that this creature in walking carried the pointed frontal part of its head shield, on which also the lateral eyes are found, raised high above the ground.[7]

In Pterygotus the four pairs of walking legs are simpler than in any other genus. They are of equal length, thin and nonspiniferous and were clearly for walking only, the prehensile function having been entirely transferred to the chelicerae.

It is manifest that in the genera in which the chelicerae are very small, especially in Eurypterus, the walking legs with their long curved spines were, as in Limulus, also actively engaged in grasping prey and transferring it to the chelicerae which transmitted it for mastication to the basal segments of the legs. There is other evidence of their prehensile function. In Limulus the first pair of walking legs of the male becomes transformed at maturity into a hooklike grasping organ and in Eurypterus fischeri, the second pair of legs also develops in the mature male a hooked clasping organ.

The first pair of postoral appendages probably also served in most genera as a tactile organ. This is very clearly indicated in Slimonia where it is directly developed into an antenniform appendage.[8] Its small size in Eusarcus and Stylonurus which contrasts with that of the following legs, is also evidence that it could have aided little, if at all, in walking or swimming.

The fifth pair of postoral appendages have been termed the ectognaths, ectognathites or swimming legs, because in most genera (Eurypterus, Dolichopterus, Eusarcus, Hughmilleria, Slimonia, Pterygotus) the terminal segments are flattened into a paddle-shaped organ that is currently considered as having functioned in swimming. Hall figured the swimming legs of the crab Platyonichus ocellatus [1861, pl. 84A, fig. 6, 7] to point out the remarkable analogy in its structure with that of the last leg of Eurypterus; and Holm has carefully worked out the characters which so excellently adapted this organ for a swimming function [1899, p. 27]. The most important of these are the sharp, knifelike edge of the anterior margins of the fourth to sixth segments, the thin blade of the seventh and eighth segments and their articulation, by which they were enabled to form a continuous oar blade at the time of the backward stroke, while in the forward stroke, the eighth segment could be turned backward on the seventh like the blade of a shears to diminish the resistance of the water. The oar blade form of the extended seventh and eighth segments is well shown on plate 4, figure 3; the reflexing of the eighth segment is observable in many other specimens [as pl. 7, fig. 7]. The oar blade is entirely smooth, without scales or hairs. The turning of this oar into a vertical position in the backward stroke probably took place mainly between the sixth and seventh segments.[9]

The material before us is competent to throw some very interesting light on the development of this swimming leg. Where typically developed, as in Eurypterus, it consists of the large basal segment, which is

Figure 11 Ventral surface of male specimen of Callinectes hastatus
ab, abdomen; pl-pv, the five pairs of legs (pereiopods); S III-VIII, sterna of thorax. (From Brooks)

followed by two ringlike segments, a longer subtubular segment (the fourth), and two shorter segments with triangular section, flat underside, sharply keeled anterior edge and expanded distally. These segments were especially active in aiding the movement of the following seventh and eighth segments which form the oar blade. The seventh segment is a flat trapezoidal plate possessing on the posterior side a triangular lobe, that served as a guard to the eighth. The latter is oval and has a terminal notch, in which a minute claw is inserted that represents a ninth segment and was termed the palette by Hall, while later authors have more properly applied the term palette to the entire eighth segment.

The species which seem to us of especial interest in the explanation of the strange structure of the swimming leg are Eurypterus (Onychopterus) kokomoensis and Dolichopterus macrochirus. The former [pl. 26, fig. 2] shows the most primitive shape of the swimming leg known to us. This primitive character manifests itself most distinctly in the following features. Beginning at the distal end, the ninth segment is a well developed terminal claw, such as is found on the other legs, indicating that the minute ninth segment of Eurypterus is a reduced terminal claw. The seventh and eighth segments do not yet so exactly fit together into a single oar blade, the seventh being still narrower and the eighth more expanded. The triangular guard lobe of the seventh segment in Eurypterus is here represented by a long relatively narrow lobe, indicating that it originated from a broadened and flattened spine of the seventh segment. A glance at figure 26, plate 2, will show that the preceding joints also are still much more uniform in character and like those of the walking legs in Eurypterus. In Dolichopterus macrochirus [see restoration, pl. 40] on the other hand, the ninth segment has been developed into a third element of the oar blade which thereby has become still more powerful. Here we find triangular guard plates on both the preceding, the seventh and eighth segments, which by their form still distinctly indicate their origin from spines. The lobelike projecting anterior portions of the distal edges of the seventh and eighth segmentsalso suggest a similar origin, especially if we compare them with the lobelike spines of the fourth pair of postoral limbs, already noted.

The postoral limbs performed a still further function, viz, mastication which rests in the basal segments termed coxae. In correspondence with this activity these are narrow, elongate, subtriangular, armed with rows of teeth borne on the narrow end (gnathobase or mandible) and form a manducatory edge. They increase in length with each successive pair and overlap like the tiles of a roof from the front backward seen in ventral view, and thus the coxae of the last pair of legs are not covered. Their sides are furnished, with smooth gliding faces. In Eurypterus and Hughmilleria the fourth coxa possesses a circular perforation covered by a thin membrane and this is also present in Limulus. Holm first observed this, suggesting from its structure and position near the inner margin of the coxa which is exposed to the outside, that it was an auditory organ. This perforation is here figured in Eurypterus remipes [pl. 7, fig. 6] and Hughmilleria socialis [pl. 62, fig. 5].

Patten [1894] has indicated that the spines on the anterior portion of the gnathobase of Limulus serve as gustatory organs. The inference is proper that the thick, blunt, hollow spines observable in like position in Eurypterus fischeri [see Holm, op. cit. pl. 2, fig. 5–8] and other eurypterids [pl. 16, fig. 1] had a like function.


Figure 12 Eurypterus fischeri Eichwald. Coxa of fourth left endognathite, seen from below (outside), showing epicoxite at right and circular perforation. (From Holm)
The coxae of the second to fourth pairs of limbs of Limulus bear a small, movable appendage behind the inner end of the manducatory edge, which is also found in the scorpion. This is known as the epicoxite. It is furnished with fine bristles and small brushes and has a tactile function. Laurie discovered this articulated process in Slimonia [1893, p. 511] but was unable to determine how many and which pairs of legs bear it. Holm found epicoxites on the first to fourth pairs of the walking legs of Eurypterus, but those of the first pair of somewhat diferent appearance and therefore possibly belonging to a different category of structure. We have also observed this appendage [see text fig. 12; pl. 57, fig. 3].

The coxa of the fifth pair differs in structure and size from those of the preceding legs. It corresponds in its large size to that of the entire leg which surpasses all other limbs in dimensions. It is of rhomboidal form, with a large neck on the anterior inner angle forming the gnathobase. The manducatory edge [pl. 72, fig. 2] is made up of an upper sharp cutting portion and a lower crushing portion consisting of a row of teeth which as a rule become finer posteriorly. The neck often becomes so lengthened as to give the coxa a retortlike appearance, as in Hughmilleria and Pterygotus and especially so in Dolichopterus and Eusarcus, where the necklike extension becomes as long as or longer than the rhomboidal base. This great extension of the gnathobase in the latter two genera is clearly correlated to the great longitudinal extension of the cephalothorax and the forward position of the mouth. In Hughmilleria, Pterygotus and especially Slimonia, the similarity to a retort is still much increased by the rounding outline of the body of the coxa. In Dolichopterus and Eusarcus the immense last coxae cover more than half of the ventral side of the cephalothorax [pl. 44]. Their form is quite characteristic in the different genera and the generic relations of detached coxae are readily recognized. While the large last coxa covers the preceding coxal segment as well as the anterior portion of the first ventral abdominal segment or operculum, the inner portion, except the chewing edge, is in turn covered by the underlip or metastoma.

The mouth, which is situated at about the middle of the ventral side of the carapace, is surrounded not only with the coxae but also by several covering liplike plates. One of these is the epistoma of Pterygotus. This was first described and figured by Huxley and Salter [Monogr. pl. 1, fig. 1]. They were, however, misled, probably by the direction of the sculpture on it, and figured it with the straight margin toward the front. Schmidt [1883, p. 71] described it as a subquadrangular plate with convex anterior and concave posterior margin, lying directly in front of the mouth and a part of the doublure of the carapace, separated from the remainder by two sutures. Laurie [1893, p. 516] has shown that the scale markings on the epistoma have their convex side directed forward contrary to the almost universal rule among eurypterids and that this fact would seem to indicate that we have here a portion of the carapace bent over. At the same time he remarks that some of his specimens are fractured along quite different lines than those Schmidt observed. One of these he reproduces on plate 2, figure 10. On another [his pl. 1, fig. 4] the "epistoma" of Slimonia is reproduced. This, however, is but the ventral marginal plate of the cephalothorax that occupies the space between the doublure of the carapace and the thin membrane surrounding the coxae and which is termed the "Randschild" by Holm. This plate in Eurypterus had been figured by Hall [pl. 80A, fig. 12] as "the lower surface of one side of the cephalic shield" and has been more fully described above. We reproduce in plate 74, figure 3 a well preserved upper lip or epistoma of P. macrophthalmus.

The term endostoma has been applied by Holm [op. cit. p. 28] to a small plate that bounds the posterior portion of the mouth. It is here figured from Pterygotus buffaloensis [plate 81, figure 4.]

It corresponds to the promesosternite of Limulus or of the scorpion group. To its anterior edge a thinner membrane is attached which passes inward in the direction of the throat and forms, therefore, the lower lip.

The metastoma or postoral plate is a highly characteristic organ of the eurypterids. It is large, somewhat variable in form but all its variations are derivable from the oval form seen in Eurypterus. Its size corresponds to the longitudinal extension of the last pair of coxae since it covers their inner margins and the interspace between them. Its frontal margin is always more or less emarginate, its margin bent under into a broad doublure connected with the membrane covering the interspaces of the ventral side. Holm has shown that there are traces still present of a bisegmented structure in the metastoma, indicating that it originated from a paired organ. He considers it homologous to the chilaria of Limulus, a pair of movable sclerites set behind the coxal segments of the last pair of legs [text fig. 10] and remarks that the metastoma of the eurypterids certainly represents a much higher development of the organ than the chilaria of Limulus. Pocock [1901, p. 302] considers the metastoma as the homologue of the sternum of the scorpion but the observations of Kishinouye [1891] upon the embryo of Limulus longispina and those of Brauer [1895] on the embryo of the scorpion demonstrate that it represents the appendages of a distinct suppressed segment. For practical reasons we have not counted this abortive first segment of the preabdomen [see diagram p. 24].

Figure 13 Eurypterus fischeri Eichwald. Endostoma, seen from below (outside). (From Holm) Figure 14 Eurypterus fischeri Eichwald. On the left the right coxa, seen from the interior and showing the doublure, the large cutting tooth and the smaller teeth; and its connection with the metastoma (on the right), which also shows its interior doublure. (From Holm)

Gaskell, in his lately published The Origin of Vertebrates, in order to derive a vertebrate prosomatic or oral chamber fully separated from the gill chambers, has assumed that the metastoma and the operculum of Eurypterus became fused [op. cit. p. 242 and our text fig. 16]. It is safe for us to say that we have no evidence in the eurypterids of any tendency toward the fusion of these organs and that it seems to us such a procedure would at once have seriously interfered with the movements of the creatures in several ways; one, because the line of fusion would be directly under the important articulation between the carapace and preabdomen, and another, it could not have failed to disturb the mutual independence of two organs of entirely distinct character and rhythm of movement, namely, the coxae of the swimming legs which the metastoma closely adjoins, and the operculum with its respiratory and sexual functions.

The form of the metastoma has been found by us to be highly characteristic of the genera of the eurypterids and to be a good indicator of their phylogenetic relations. That of Eurypterus is typically oval in

Figure 15 Eurypterus fischeri Eichwald. Metastoma, showing the deep furrow of the anterior portion. (From Holm)

Figure 16 Diagram of sagittal median section through Eurypterus: 7, metastoma; 8, operculum; 2, 3, 4, 5, endognathites. (Gaskell's reconstruction)

outline, while those of Hughmilleria and Pterygotus show a strong tendency to become wider in the anterior half and narrower in the posterior, and at the same time more deeply emarginate in front. In Slimonia this tendency is carried to an extreme, the metastoma having become very elongate-cordate with a narrow posterior half and a deeply emarginate anterior one. In Dolichopterus and the Stylonurus branch in general, again a different tendency is developed. Here the base of the metastoma becomes rectangular, the lateral margins subparallel and the plate approaches a long rectangle, with the short front side deeply emarginate (Stylonurus) or it becomes lyrate (Dolichopterus). Finally, in Eusarcus the metastoma has become subtriangular or cordate in outline, the anterior portion being greatly widened and emarginate and the posterior tapering to a blunt point.

We have thus at least four distinct lines of development of the metastoma, which fully correspond to the four principal branches of the eurypterids here distinguished [see chapter on Phylogeny, p. 124] namely, that of Eurypterus, of Pterygotus, of Eusarcus and of Stylonurus.

Figure 17 Metastomas. I, of Eurypterus; II, of Eusarcus; III, of Dolichopterus; IV, of Stylonurus; V, of Hughmilleria; VI, of Pterygotus; VII, of Slimonia

Abdomen. The abdomen consists of 12 segments or somites. The anterior six of these are divided into separate dorsal and ventral pieces. They form the preabdomen or mesosoma which is also sometimes termed the thorax; the six posterior ones are annular and form the postabdomen or metasoma, the abdomen or tail of earlier writers. The dorsal plates of the preabdominal or mesosomatic segments are termed the tergites, the ventral pieces the sternites or Blattfüsse. Those of the postabdomen are known as caudal, postabdominal or metasomatic segments or somites.

The preabdomen is widest at the fourth or fifth tergite whence it usually contracts more rapidly. It formed a unit in the movements of the body, the easier articulation taking place between it and the cephalothorax on one side and at the boundary of preabdomen and postabdomen on the other. The first tergite is a narrow plate curved backward and with rounded ends. The other tergites are transverse bandlike plates, with convex anterior and concave posterior margins in the middle and the lateral ends curved slightly forward. Corresponding to this outline of the plates the middle portion of the preabdomen is elevated forming the rhachis, while the wings are often depressed or concave. This lateral portion of the segment is frequently termed the epimeral portion, epimera or pleura [pl. 5, fig. 3]. The epimera are produced at the antelateral angle into lobes, or ears, especially distinct in Pterygotus. These "ears" have been considered as serving for the attachment of muscles but according to Schmidt they only correspond to the rounding of the postlateral angles and served merely to protect the outside of the body. The lateral and posterior margins are furnished with more or less broad doublures to which the connecting membrane is attached. The anterior margin of the tergites which is overlapped by the preceding one (except that of the first tergite) is smooth or bears only very fine ornamentation and is depressed, thus forming an articulation with the doublure of the preceding segment. In some genera, as Eurypterus, this articulation extends the whole width of the tergite; in others, as Pterygotus, only the rhachis is provided with a distinct articulation; this difference probably indicating different degrees of mobility of the preabdomen.

The posterior margin of the gliding or articulating face of the tergite is mostly bounded by a continuous transverse line of scales [pl. 8, fig. 2].

Corresponding to the six tergites are only five ventral plates or sternites. This is due to the fact that the first two ventral segments lack the ventral sclerites, their place being taken by the large genital plate or operculum, homologous to the operculum of Limulus which bears the generative organs. The operculum of the eurypterids consists of a pair of plates meeting in the median line and having a median lobe attached to them. The plates have a straight anterior margin and frequently well rounded antelateral angles. The scales form a continuous transverse line across the plates which has been erroneously considered as the suture resulting from the fusion of the two sternites. In front of the median lobe two triangular—or sometimes pentagonal—areas are marked off [pl. 11, fig. 3] by sutures from the opercular plates in Eurypterus and Slimonia but rarely in Pterygotus. Laurie suggests that these areas may represent the paired sternite of the first abdominal segment, the remaining portions of the plates representing the appendages.

The middle lobe shows two different forms not known to Hall but which were recorded by Woodward in Pterygotus bilobus and Slimonia acuminata [1863, p. 61; 1872, p. 114, f.] and attributed to sexual differences. Schmidt likewise recognized two forms of opercular appendages of sexual significance in Eurypterus fischeri, and Holm, by reference to Limulus, agreed with Woodward in assigning the more primitive appendage to the male and the more elaborate to the female, thus bringing out the fact that the mature males, at least in Eurypterus, are smaller than the females, as is true of Limulus. Gaskell, however, asserts [1908, p. 191] that the operculum of the eurypterids belonged to the type of Thelyphonus rather than to that of Limulus or Scorpio and as appears from his diagram [see text fig. 18] he would, on the strength of this claim, reverse the reference of the appendages to the sexes.[10] While it may be that the elaborate opercular appendages of the eurypterids exhibit less similarity to the extremely primitive exterior genital apparatus of Limulus than to that of Thelyphonus, we must not forget that Holm has found important corroborative evidence for his sex determination of the male in the clasping organs of the second endognathites.

Accepting the determinations of Woodward, Schmidt and Holm, the female appendage consists of two single lobes and two paired terminal

Figure 18 Diagram to indicate the probable nature of the mesosomatic segments of Eurypterus. Ut. masc., Uterus masculinus; gen duct, genital ducts. (From Gaskell)

Figure 19 Eurypterus fischeri Eichwald. Female genital appendage of operculum. At the right the tubular appendage. (From Holm)

Figure 20 Eurypterus lacustris Harlan. Specimen showing the paired genital appendages turned to one side. Natural size

pieces.[11] The first of the single lobes is the largest; it is pointed in front, where it extends between the pentagonal basal pieces, with which it is connected by sutures. The middle portion is found occupying the space between the lateral opercular plates which here do not meet in the median line. The distal end is produced into two short pointed pieces. The second single piece underlies or is telescopically pushed into the first and is of the same or very similar form as the first. Connected with the female genital apparatus were paired internal tubular appendages first correctly recognized in E. fischeri. These are also well seen in specimens of E. remipes [pl. 8, fig. 1] and E. lacustris [pl. 12, fig. 2]. In one of the representatives of the latter species [Buffalo Society of Natural Sciences] both tubes lie to one side of the median lobe [see text fig. 20] thereby indicating that they extended free into the interior of the body.

In the male the two lateral opercular plates are more regularly rectangular in outline and come into contact along the median line. The genital appendage is very small and composed only of two single pieces. No pentagonal areas are set off by sutures. The principal single appendage consists of a median piece with parallel sides exposed to the outside and two larger wings underlying the opercular plates; the second appendage is very small, triangular and adjoins the first.

In other genera, as Hughmilleria and Pterygotus, the genital appendages are clearly much simpler. In Hughmilleria only one sagittate-based lobe is found on the female operculum [pl. 62, fig. 9, 10] and in the male a convex, broadly lanceolate lobe [pl. 62, fig. 11].

In Pterygotus these sexual appendages are not yet clearly distinguished, but they seem to be little advanced beyond those of Hughmilleria. The female appendage in its simplest form [P. bilobus, see Woodward, 1869, p. 61, pl. 12, fig. 1C; Laurie, 1893, pl. 2, fig. 14] is a straight, narrow, very slightly expanding plate with a ridge down the middle and ending in a bluntly triangular point, while, where more highly developed, as in P. osiliensis, it possesses a rhombic sagittate base, connected by sutures with the lateral opercular plates, and continued into the free club-shaped principal part with rounded extremity. In some cases, as in P. anglicus, [see Huxley and Salter, pl. 3] and in an unidentified form from Otisville [pl. 78, fig. 3] the extremity was expanded into a disklike body, the whole appendage there resembling a pendulum. Another, supposedly the male, appendage has been figured by Woodward [1869, p. 61] and by Laurie [1893, pl. 2, fig. 13]. This is hastate at its proximal end, and sharply pointed at its distal extremity in Woodward's figure while Laurie represents it as a short, blunt process, but Laurie's figure suggests that it was taken from a fragmentary specimen. Woodward and Laurie also figure the pair of triangular areas in front of the process. Schmidt mentions only faint traces of the sutures in P. osiliensis [op. cit. p. 78] and our material has not shown them at all. As they are also absent in Hughmilleria, they are probably a new acquisition in some species of Pterygotus.

In Slimonia the appendages have retained fundamentally the same structure as in Pterygotus, but they show greater elaboration. The frontal

Figure 21 Slimonia acuminata (Salter). Opercular appendages. (From Woodward)

paired triangular areas are distinctly set off. That form of the median lobe which is considered as belonging to the female [see Woodward, 1872, p. 116, fig. 35; Laurie, 1893, p. 513] terminates in three sharp points at its free end, while the other or male form terminates in a more or less truncated cone. This shows two or three deep transverse furrows, which Laurie thinks due to its having been eversible [op. cit. pl. 2, fig. 8].

In Eusarcus the genital appendages are still incompletely known. We have seen only fragments of the female appendage [pl. 33, fig. 3; text fig. 55] which exhibit the triangular areas, the hastate basal portion of the first lobe, and a faint impression of the second lobe. Indications of the interior tubular appendages have also been seen. According to this evidence the whole organ is a simpler expression of that in Eurypterus.

A fairly well preserved genital appendage of a female Dolichopterus macrochirus [pl. 44] has been observed by us. It is clearly built on the type of that of Eurypterus in showing the two invaginated median lobes with their bifid extremities. The organ was also, as in Eurypterus, of relatively large size.

The genital appendages of Stylonurus are not known.

The, second sternite has also been found to bear a genital appendage in the females of Eurypterus, Hughmilleria and Dolichopterus. In Eurypterus it consists of one short unpaired proximal and two long awl-shaped distal pieces, the whole being covered by the opercular appendage. In Dolichopterus we have seen the impression of two similar slender terminal pieces between the shorter hornlike terminal pieces of the opercular appendage, and therefore feel sure that it possessed an appendage of the second sternite like that of Eurypterus. In Hughmilleria the appendage of the second sternite has been fully described and figured by Sarle [op. cit. pl. 62, fig. 9, 10]. It is small and attenuate with a triangular base.

Probably the second sternite of the female was also furnished with a small appendage in some of the other genera where it has thus far escaped observation.

From present evidence it may be stated as a general proposition that there are two lines of development of the genital appendages of the operculum in the eurypterids, those of the Pterygotus group and those of the Eurypterus group. In the first, the appendage is composed of but one unpaired lobe, which may become more or less elaborated, as in Slimonia. In the second group two unpaired and two paired lobes are developed.

There is little known as to the genital openings in the eurypterids. Woodward has figured [1872, p. 117, fig. 36] small openings on round tubercles on the basal triangular areas of the opercular appendages in Slimonia, which he considers as ovarian openings. Laurie, however, has not mentioned or figured these openings in his investigation of Slimonia [1893] and their presence seems open to doubt, especially since no homologous apertures have been seen in the other genera. Neither have other authors taken up this problem as far as we are aware. In several species we have found distinct openings in the corners where the posterior points of the triangular areas of the opercular plates meet the lateral points of the hastate basal portion in the unpaired lobe of the female genital appendage. These lateral points are as a rule extended into earlike or scroll-like lobes that cover the openings. Since the paired interior tubular appendages observed in Eurypterus end at these lateral points, it is certain that they emptied here, as suggested by Holm, who considers them as auxiliary generative organs.

The large opercular appendage of the female is but a median tongue and not a tube, and it is probable that it concealed the complicated terminal portions of the genital organs, as the median part of the operculum still does today in the arachnids Phrynus and Thelyphonus. Gaskell [p. 192] has drawn a restoration, representing these genital organs "in accordance with our knowledge of the nature of these organs in the present-day scorpions, as a median elongated uterus, bilaterally formed, from which the genital ducts passed, probably as in Limulus, towards a mass of generative glands in the cephalic region, and not as in Scorpio or Thelyphonus, tailwards to the abdominal region." We surmise that the female eurypterid thrust this chitinous opercular appendage, together with the opercular plate, into the sand during oviposition, just as Limulus does today.

It is possible that the paired tubular organs mentioned above are not auxiliary generative organs, as surmised by Holm, but are the genital ducts and that the openings observed by the writers are the apertures of these ducts leading into the median uterus.

The sternal plates of the third to the sixth mesosomatic somites are well developed, forming the four sternites following the operculum. These plates are always slightly arched forward, the anterior margin being slightly convex, the posterior concave. The antelateral angles are produced into rounded lobes and the median line is marked by a suture, along which they readily separated. The sternites were often more arched than the tergites, as shown by uncompressed specimens [pl. 20, fig. 8] and by the common observation that in compressed specimens the sternites protrude on both sides from under the tergites [pl. 4, fig. 4]. Although there are but five sternites including the operculum corresponding to the six tergites of the preabdomen the former are relatively so much longer that they overlap each other with fully half their length.


Figure 22 Eurypterus fischeri Eichwald. Portion of one of the posterior sternites, showing anteriorly the very delicate membrane of the interior side torn off and pushed forward and exhibiting the oval attachment area of the gills. ×3. (From Holm)
The sternites, the operculum included, bore the branchiae. Woodward was the first to observe platelike appendages of the sternites in Pterygotus which he described as lamellae, and Laurie observed appendages of the sternites in Slimonia which he designated as "branchial lamellae." Holm found organs in Eurypterus corresponding to those described by Laurie, but considers them only as "Kiemenplatten" instead of "Kiemenblätter," or as oval spongy thickenings of the outside of the thin, soft membrane on the upper side of the sternite, which probably served as attachment places to the branchial lamellae. He has also observed detached bundles of two or three extremely thin superjacent leaves which he considers with doubt as possible branchial lamellae. The "Kiemenplatte" or "branchial plate" exhibits a very characteristic structure consisting of one or two trunk veins running parallel to the longitudinal extension of the plate and from which smaller branches proceed [see text fig. 22].

In our material we have frequently been able to see the impression of the "branchial plates" from the dorsal side, as in the fine specimen of Eusarcus [pl. 29]. Detached branchial plates have also been observed and in the remarkable specimen of Eurypterus kokomoensis, reproduced in plate 26, figure 2, the branchial plates are distinctly set off from a very tenuous brownish film, the integument of the individual, as thick, jet-black oval plates with a coarsely granular surface.[12] These gill plates were here very much smaller than those of the following sternites, those of the third and fourth sternites being the largest and those of the last pair of sternites again considerably smaller. On the plates of the operculum and of the next sternite, projecting thick-walled tubes can be seen that perhaps correspond to the trunk veins observed by Holm.

Metasoma, postabdomen or tail. The tail of the eurypterids consists of six ringlike segments, which decrease in width and correspondingly increase in length in posterior direction. In the primitive genus Strabops they merely decrease in width; and in all more primitive forms, as Strabops and Hughmilleria, and the simpler species of Eurypterus, the decrease in width is very gradual. In others it is abrupt in the first and second postabdominal segments. The first postabdominal segment in most species is a strongly contracted conical ring. In forms where the preabdomen has been excessively broad, the contraction takes place mainly in the last two tergites and the first postabdominal segment. In Eusarcus, where the contrast between the broad preabdomen and the narrow taillike postabdomen has become extreme, the first caudal segment contracts by more than one half, while the following segments have nearly uniform width, forming a cylindrical tail.

The segments fit into each other like the joints of a telescope, the posterior one always reaching, with its anterior articulating edge, as far as the posterior doublure of the anterior segment. They consequently articulate in all directions and are therefore found either extended straight or curved to either side. An exception to this rule seems to be presented by Eusarcus in which the tail is nearly always curved to one side— although specimens with straight tails have been observed in the shorttailed E. newlini; and moreover is so twisted that while the first caudal segments show the dorsal or ventral side, the last and the telson are seen in profile. This fact, together with the presence of an emargination in the posterior upper margin of the segments and the downward curvature, seems to indicate that the tail was capable of bending upward and forward as in the scorpion.

The section of the postabdomen usually was lenticular at the anterior end, remaining so throughout in some forms, as some species of Eurypterus, while in many species it became circular toward the other end. Circular sections of the caudal segments of species of Eurypterus and Hughmilleria are frequently found [pl. 63, fig. 12].

The lateral angles of the posterior margins are nearly always produced into pointed lobes. Those of the ultimate segments are usually much larger than those of the preceding segments, as e. g. in E. dekayi. In some species these lobes, together with the lateral edges of the segments, grow out into prominent flat winglike appendages, representing the "epimeral pieces" of the caudal segments. These are notably developed in Dolichopterus macrochirus, in the subgenus Anthraconectes and in some species of Stylonurus.

The ultimate and penultimate segments of Pterygotus carry on the dorsal side a crest or ridge that begins at about the middle and continues to the posterior end. It is doubtless caused by the intestinal canal.

The telson is an appendage of the 12th segment, as indicated by the position of the anus in relation to it. Strabops and our larval stages seem to indicate that the primitive form of this spine was short, thick and four-sided with dorsal, ventral and two lateral edges. This view is supported by the consideration that from this pyramidal form we can most easily derive the two different lines of development of the telson, which culminated on one hand in the styliform telson of Stylonurus and on the other in the bilobed telson of Erettopterus. The more primitive genera, such as Eurypterus and Hughmilleria, retain the general character of the primitive telson and at the same time exhibit in an incipient stage the new features that lead to Stylonurus and Erettopterus. In Eurypterus also the telson has still the original four-sided form, but the two upper sides have already become so reduced that they are united in either a flat or a concave broad dorsal side, while the ventral edge is developed into a flat-topped carina.[13] In Hughmilleria the development is reversed, the dorsal side bearing a carina, and the ventral side being smooth and either flat or but slightly convex, the telson having thus the subtriangular section and dorsal carina of that in Limulus. The essential characters of the telson of Eurypterus are retained in Dolichopterus. In Anthraconectes, however, it becomes extremely long and styliform, thus assuming the characters in Drepanopterus and Stylonurus. In the extreme forms of the latter genera the telson becomes contracted in the proximal portion and expanded clublike in the distal portion, sometimes with development of flat lateral carinae or flanges, as in Stylonurus scoticus.

The telson of Eusarcus was apparently still four-keeled; but it was bent downward toward the end, so that when the scorpionlike tail was thrown forward over the body, its sharp point would be directed upward. This is an aberrant development of the telson not found in other genera, but clearly connected with that of Eurypterus.

From the telson of Hughmilleria first that of Slimonia, and through the latter, those of Pterygotus and Erettopterus can be derived. Slimonia possesses a long tail spine with dorsal carination. The anterior portions of the lateral carinae of this spine, however, are developed into two broad flanges which together form an oval, leaf like expansion with coarsely serrate posterior margin, the median keel being continuous with the long spine [see text fig. 25]. In Pterygotus the projecting spine of the Slimonia stage has been reduced to the size of the serrations of the flanges, and in Erettopterus the reduction of the median spine that has become the axis of the broad leaflike telson has been carried still further so that a bilobed telson has resulted.

The keeled lateral and ventral edges of the tail spine in Eurypterus and the lateral edges of the broad telson in Pterygotus and Erettopterus are, as a rule, furnished with serrations, which become more prominent in posterior direction.

The articulation of the telson is quite like that of Limulus, consisting of a broad transverse lower and a small upper segment and is mainly adapted to movement in a vertical plane.


  1. The muscles of the eurypterid cephalothorax were probably contracted into solid bundles, as in Limulus, following the solidification of the segments. This is indicated by the occurrence of a pair of entapophyses, or chitinous infoldings of the carapace [pl. 6, fig. 6],
    Figure 4 Dorso-ventral muscles on carapace of Eurypterus
    le, lateral eyes; ce, ocelli; 62, median dorsopreoral-entosclerite muscle; 63, anterior dorso-plastron muscles; 64, median dorso-plastron muscle. (From Gaskell)
    corresponding in their position to those observed in Limulus, and also by the observation of pairs of circular or oval areas arranged along the median line. At least three such pairs of areas have been seen in Eurypterus remipes. The foremost of them is shown in plate 5, figure 7. They may correspond to the first pair of tergo-proplastrals in Limulus which suspend the anterior horns of the entosternon from the dorsal side of the carapace. Gaskell has drawn a hypothetical restoration of the muscle system of the cephalothorax of an Eurypterus [text fig. 4] by inserting in the carapace of E. scouleri the segmental dorso-ventral muscles as met with in the living scorpion. In a general way this is corroborated by our observations, with the exception that the muscles do not appear to be arranged so closely around the center of the carapace, at least in Eurypterus remipes. From evidence presented subsequently it would seem preferable to use Limulus as the basis of such restoration.
  2. The form of the carapace varies with the position of the eyes, the form of the legs and the mode of life of the animals. We shall note these relations more fully in another chapter.
  3. The importance of this internal skeleton to the muscular system of Limulus is fully described by Owen [Palaeontographical Society 1878. 32:187].
  4. It may be noted in this place that a specimen in Buffalo exhibits indications of a faint apical depression of the papillae.
  5. See Parker, G. H., The Compound Eyes in Crustaceans. Mus. Comp. Zool. Bul 1891. v. 21.
  6. Our common crab (Callinectes hastatus) also appears to use the fourth pair of legs, that in front of the paddle-shaped swimming legs, as a kind of balancer in swimming.
  7. We have described more fully in the generic discussion the peculiar character of the genus as expressed in its appendages.
  8. The following pairs of walking legs also show a peculiar development in Slimonia in being equally as long as in Pterygotus, but having a fringe of long spines on the distal edges of the segments, evidently developed from the serrations seen in Eurypterus and other genera, while the paired spines are absent.
  9. Laurie [1893, p. 124] has suggested that "it seems more probable, that the foot was used for anchoring the animal firmly in the soft mud of the sea bottom, and possibly also for shoveling up the sand and mud when the beast wished to conceal itself." We shall recur to the question of the use of the last pair of legs in our further discussion.
  10. In this connection Gaskell has introduced [op. cit. fig. 78, p. 191] a figure which is stated to "be a picture from Schmidt of the ventral aspect of Eurypterus," but in fact is a monstrous mixture of the characters of Eurypterus and Pterygotus, such as Friedr. Schmidt could not possibly have perpetrated even in a nightmare. This is a good example of the careless treatment of fossils in zoological textbooks and treatises, exemplified again by the figure of a Pterygotus anglicus labeled as "Eurypterus remipes" in the Text-book of Invertebrate Morphology, by J. P. McMurrich.
  11. Holm also counts the two pentagonal anterior areas [see above] as belonging to the genital appendage.
  12. The plates are much more distinct on the specimen than the photograph was able to bring out, owing to a thin whitish film over them.
  13. It is possible that this section is but the result of compression and desiccation and that the living specimens had a trapezoidal section.