have been domesticated by these savages, and the condition of the bones themselves confirms the idea that they belonged to wild animals. The sheep and the reindeer being entirely absent, and the domestic cat not having been known in Europe until about the ninth century, the dog appears to have been the only domestic animal of the period; and though it may fairly be asked whether the bones may not have belonged to a race of wild dogs, the question admits of a satisfactory answer.
Among the remains of birds, the long bones which form about one-fifth of the skeleton are, in the Kjökkenmöddings, about twenty times as numerous as the others, and are almost always imperfect, the shaft only remaining. In the same manner it would be impossible to reconstruct a perfect skeleton of the quadrupeds, certain bones and parts of bones being always absent. In the case of the ox, for instance, the missing parts are the heads of the long bones (though while the shaft only of the femur is found, in the humerus one end is generally perfect), the back bone except the two first vertebræ, the spinous processes, and generally the ribs, and the bones of the skull except the lower jaw and the portion round the eyes. It occurred to M. Steenstrup that these curious facts might, perhaps, be referred to the dogs; and, on trying the experiment, he ascertained that the bones which are absent from the Kjökkenmöddings are precisely those which the dogs eat, and those which are present are the parts which are too hard and solid to contain much nourishment. M. Steenstrup called my attention to a diagram of a bird's skeleton, tinted in such a manner as to show at a glance which of the bones occur in the Kjökkenmöddings, and pointed out to me that it coincided exactly with one given by M. Serres to illustrate those portions of the skeleton which were first formed.
Although a glance at, for instance, a femur, and a comparison of the open cancellated tissue of the two ends with the solid, close, texture of the shaft, at once justifies and accounts for the selection made by the dogs, it is interesting thus to ascertain that their predilections were the same in primæval times as at present. Moreover, we may in this manner explain the prevalence of some bones in fossil strata, I have already mentioned that of the skull, the hard parts round the eye and the lower jaw are the only parts left; now, the preponderance of lower jaws in a fossil state is well known.
In the "Proceedings of the Geological Society for 1857," p. 277, Dr. Falconer, after describing some of the fossils found by Mr. Beccles at Swanage, says:—"The curious fact that only lower jaws should have turned up among the Stonesfield mammalian remains has often been the subject of speculation or remark. The same, to a certain extent, has held good with the remains found in the Purbeck beds. . . . . . In these minute creatures, unless the bone be complete, and, supposing it to be a long bone, with both its articular surfaces perfect, it is almost hopeless, or at any rate very discouraging, to attempt to make out the creature that yielded it; whereas the smallest fragment of a jaw, with a minute tooth in it,
