ferent type from those found at the sandhills at the mouth of the Shag River associated with skulls with rounded beaks. This larger species of elephantopus forms the type of Lydekker's genus Pachyornis, and the genus Euryapteryx will contain only the blunt-beaked species E. ponderosa and E. gravis. I find also from the Shag Point collection that the sternum of Euryapteryx resembles that of Mesopteryx (or Syornis), but is more robust and has no pneumatic depressions at the inner anterior corners. An examination of the collection from Enfield has shown me that rheïdes and casuarina are congeneric with didina, and will all come into the genus Mesopteryx, distinguished by its slender pointed beak and shallow temporal fossæ. This leaves crassus isolated; and, for the present, it must take Reichenbach's generic name of Emeus. The skeleton of Emeus is, however, very imperfectly known. A skull from Enfield, which I refer to it, comes near Mesopteryx, but has narrower and deeper temporal fossæ. The sternum which I refer provisionally to Emeus resembles that of Euryapteryx, but has slight pneumatic depressions. In fact, Emeus appears to be intermediate between Mesopteryx and Euryapteryx, and further information is necessary before we can say whether it should be joined to either of them. It may be thought by some that too many species of moas have already been made; but with this I cannot agree. The study of individual skeletons has shown me that bones, which I formerly regarded as being merely varieties of one species, belong really to different genera; and the more I study the moas the more I see the necessity of limiting the amount of variation allowed to each species. An important biological problem, perhaps not found elsewhere, is presented to us in the development of the moas. This is the variation of an herbivorous race of animals, well supplied with food, and without the check due to the presence of carnivorous mammals preying on them. This problem can only be attacked by studying closely the distribution of the moas both in time and in space; and to do so it is necessary that the species should be closely defined. Again, the relative ages of our Pliocene and Pleistocene terrestrial deposits can probably only be ascertained by the help of moa bones, and here, again, unless the different species and varieties are kept separate, we can never obtain the data necessary to settle these questions. It may, however, be said, Why give them all names? Why not consider them as varieties of a few species? The answer is that there is still plenty of room for varieties within the limits of these species. For example, individuals of Dinornis robustus and D. strenuus from Enfield are larger than those from Hamilton. The specimens of D. torosus from Enfield