590 NATURAL SCIENCE. oct. 



level, similar to those just mentioned. In the Miocene period it 

 consisted of a cluster of several islands, which were elevated and 

 united in the Older Pliocene, and ultimately divided into the two 

 islands we have now in the Newer Pliocene. If the ancestors of the 

 Moas inhabited New Zealand during the Eocene period, they must 

 have been separated on these islands during the whole of the 

 Miocene, and mingled together again in the Pliocene. In this way — 

 i.e., by isolation — probably the genera originated, but the species 

 appear to be due to variations without isolation. As is the case with 

 most common animals, the Moas varied greatly, and there being no 

 carnivorous animals to hold them in check, while vegetable food was 

 abundant, natural selection did not come into play, and the inter- 

 mediate forms were not strictly eliminated. Under such favourable 

 circumstances conditions of life were easy, and the birds got larger 

 and fatter, more sluggish and more stupid. The oldest known Moa 

 is one of the smallest, and it is the smaller species which are found 

 in both islands ; from which we may infer that they were the only 

 ones in existence when the two islands were united, and that the 

 Moas since then increased in size. But the very large Moas were 

 always comparatively rare. The commonest kinds in the North 

 Island were only from two-and-a-half to four feet high, while those of 

 the South Island were mostly from four to six feet in height. The 

 giant forms, going up twelve and thirteen feet, were seldom seen. 



Throughout the Pliocene period the Moas flourished greatly ; 

 but in the Pleistocene they must, in the South Island, have died in 

 large numbers, for how else could such immense quantities of bones 

 have got together in the peat-beds at Glenmark and at Hamilton in 

 Central Otago ? It has often been suggested that flocks of birds, 

 attempting to escape from fires, rushed into the swamps and perished. 

 But when we remember that these Moas died thousands of years ago, 

 long before there were any human inhabitants to light fires, it will be 

 seen that this surmise is quite out of the question. Only two hypo- 

 theses appear to be possible to account for the facts. Either the 

 birds walked into the swamp and were drowned, or else their dead 

 bodies were washed in. The first hypothesis is probably the expla- 

 nation of the deposit at Te Ante near Napier, because many of the 

 leg bones were found upright in their natural position. But at 

 Glenmark and at Hamilton the bones were lying in all directions, as 

 often upside down as in any other position, and the peat-beds were 

 only a few feet thick, and filled with bones up to the very top. We 

 cannot, therefore, suppose that these Moas were swamped, and there 

 is evidence in both of these cases to show that the dead bodies of 

 birds were washed in by floods. We find corroborative evidence of 

 this in the alluvial plains of Central Otago, for these always contain 

 numerous bones wherever a stream enters them from the hills. 



But how are we to account for the number of dead birds washed 

 down from the hills? There are two remarkable facts connected 



