of the somites, and, for the most part, of the limbs. At the same time, the tendency to a retardation in the development of the posterior thoracic somites is very general in Malacostracan larvae, and may perhaps be correlated with the fact that in the primitive Phyllocarida the whole thoracic region is very short and the limbs closely crowded together.
 |
| Fig. 15.—Nauplius of Tetraclita
porosa after the first moult. |
| (Fritz Müller.) |
Besides the nauplius and the zoea there are many other types of Crustacean larvae, distinguished by special names, though, as their occurrence is restricted within the limits of the smaller systematic groups, they are of less general interest. We need only mention the Mysis-stage (better termed Schizopod-stage) found in many Macrura (as, for example, the lobster), which differs from the adult in having large natatory exopodites on the thoracic legs.
Most of the larval forms swim freely at the surface of the sea, and many show special adaptations to this habit of life. As in many other “pelagic” organisms, spines and processes from the surface of the body are often developed, which are probably less important as defensive organs than as aids to flotation. This is well seen in the nauplius of many Cirripedia (fig. 15) and in nearly all zoeae. Perhaps the most striking example is the zoea-like larva of the Sergestidae, known as Elaphocaris, which has an extraordinary armature of ramified spines. The same purpose is probably served by the extreme flattening of the body in the membranous Phyllosoma-larva of the rock-lobsters and their allies (Loricata).
Although fossil remains of Crustacea are abundant, from the most ancient fossiliferous rocks down to the most recent, their study has hitherto contributed little to a precise knowledge of the phylogenetic history of the class. This is partly due to the fact that many important forms must have escaped fossilization altogether owing to their small size and delicate structure, while very many of those actually preserved are known only from the carapace or shell, the limbs being absent or represented only by indecipherable fragments. Further, many important groups were already differentiated when the geological record began. The Phyllopoda, Ostracoda and Cirripedia (Thyrostraca) are represented in Cambrian or Silurian rocks by forms which seem to have resembled closely those now existing, so that palaeontology can have little light to throw on the mode of origin of these groups. With the Malacostraca the case is little better. There is considerable reason for believing that the Ceratiocaridae, which are found from the Cambrian onwards, were allied to the existing Nebalia, and may possibly include the forerunners of the true Malacostraca, but nothing is definitely known of their appendages. In Palaeozoic formations, from the Upper Devonian onwards, numbers of shrimp-like forms are found which have been referred to the Schizopoda and the Decapoda, but here again the scanty information which may be gleaned as to the structure of the limbs rarely permits of definite conclusions as to their affinities. The recent discovery in the Tasmanian “schizopod” Anaspides, of what is believed to be a living representative of the Carboniferous and Permian Syncarida, has, however, afforded a clue to the affinities of some of these problematical forms.
True Decapods are first met with in Mesozoic rocks, the first to appear being the Penaeidea, a primitive group comprising the Penaeidae and Sergestidae, which occur in the Jurassic and perhaps in the Trias. Some of the earliest are referred to the existing genus Penaeus. The Stenopidea, another primitive group, differing from the Penaeidea in the character of the gills, appear in the Trias and Jurassic. The Caridea or true prawns and shrimps appear later, in the Upper Jurassic, some of them presenting primitive characteristics in the retention of swimming exopodites on the walking-legs. The Eryonidea (fig. 16, 3), a group related to the Loricata but of a more generalized type, are specially interesting since the few existing deep-sea forms appear to be only surviving remnants of what was, in the Mesozoic period, a dominant group. The Mesozoic Glyphaeidae have been supposed to stand in the direct line of descent of the modern rock-lobsters and their allies (Loricata). Some of the Loricata have persisted with little change from the Cretaceous period to the present day.
The Anomura are hardly known as fossils. The Brachyura, on the other hand, are well represented (fig 16, 1, 2). The earliest forms, from the Lower Oolite and later, belonging chiefly to the extinct family Prosoponidae, have been shown to have close relations with the most generalized of existing Brachyura, the deep-sea Homolodromiidae, and to link the Brachyura to the Homarine (lobster-like) Macrura.
A few Isopoda are known from Secondary rocks, but their systematic position is doubtful and they throw no light on the evolution of the group. The Amphipoda are not definitely known to occur till Tertiary times. Stomatopoda of a very modern-looking type, and even their larvae, occur in Jurassic rocks.
 | |
| Fig. 16. | |
| 1, Dromilites Lamarckii, Desm.; London Clay, Sheppey. 2, Palaeocorystes Stokesii, Gault; Folkestone. 3, Eryon arctiformis, Schl.; Lithographic stone, Solenhofen. |
4, Mecocheirus longimanus, Schl.; Lithographic stone, Solenhofen. 5, Cypridea tuberculata, Sby.; (Ostracoda); Weald, Sussex. 6, Loricula pulchella, Sby (Cirri- pedia); L. Chalk, Sussex. |
In the dearth of trustworthy evidence as to the actual forerunners of existing Crustacea, we are compelled to rely wholly on the data afforded by comparative anatomy and embryology in attempting to reconstruct the probable phylogeny of the class. It is unnecessary to insist on the purely speculative character of the conclusions to be reached in this way, so long as they cannot be checked by the results of palaeontology, but, when this is recognized, such speculation is not only legitimate but necessary as a basis on which to build a natural classification.
The first attempts to reconstruct the genealogical history of the Crustacea started from the assumption that the “theory of recapitulation” could be applied to their larval history. The various larval forms, especially the nauplius and zoea, were supposed to reproduce, more or less closely, the actual structure of ancestral types. So far as the zoea was concerned, this assumption was soon shown to be erroneous, and the secondary nature of this type of larva is now generally admitted. As regards the nauplius, however, the constancy of its general character in the most widely diverse groups of Crustacea strongly suggests that it is a very ancient type, and the view has been advocated that the Crustacea must have arisen from an unsegmented nauplius-like ancestor.
The objections to this view, however, are considerable. The resemblances between the Crustacea and the Annelid worms, in such characters as the structure of the nervous system and the mode of growth of the somites, can hardly be ignored. Several structures which must be attributed, to the common
