680 



SCIENCE. 



[N. S. Vol. IV. No. 97. 



from wliich they almost certainly arose, 

 the Bullidse, are not known before the 

 Trias. Furthermore, the forms which are 

 clearly the oldest of the Pteropods — Lima- 

 cina and Spirialis — are not known before 

 the beginning of the Tertiary Period. 

 Either there is a mistake in the identifica- 

 tion of the Palaeozoic fossils as Pteropods, 

 or the record is even more incomplete than 

 usual, and the most specialized of all Mol- 

 luscan groups had been formed before the 

 date of the earliest fossiliferous rocks. If 

 this should hereafter be disproved, there 

 can be no doubt about the early appear- 

 ance of the MoUuscan classes, and that 

 it is the irony of an incomplete record 

 which places the Cephalopods and Gastro- 

 pods in the Cambrian and the far more an- 

 cestral Chiton no lower than the Silurian. 

 Throughout the fossiliferous series the older 

 families of Gastropods and Lamellibranchs 

 are followed by numerous other families, 

 which were doubtless derived from them ; 

 new and higher groups of Cephalopods were 

 developed, and, with the older groups, either 

 persisted until the present time or became 

 extinct. But in all this splitting up of the 

 classes into groups of not widely different 

 morphological value, there is very little 

 progressive modification, and, taking such 

 changes in such a period as our unit for the 

 determination of the time which was neces- 

 sary for the origin of the classes from a 

 form like Chiton, we are led to the same 

 conclusion as that which followed from the 

 consideration of the Appendiculata, viz : 

 that the fossiliferous series would have to 

 be multiplied several times in order to pro- 

 vide it. 



Of the Phylum Gephyrea, I will only 

 mention the Brachiopods, which are found 

 in immense profusion in the early Palaeo- 

 zoic rocks and which have occupied the 

 subsequent time in becoming less dominant 

 and important. So far from helping us to 

 clear up the mystery which surrounds the 



origin of the class, the earliest forms are 

 quite as specialized as those living now, 

 and, some of them (Lingula, Discina) even 

 generically identical. The demand for time 

 to originate the group is quite as grasping 

 as that of the others we have been con- 

 sidering. 



All the classes of Echinoderma, except 

 the Holothurians, which do not possess a 

 structure favorable for fossilization, are 

 found early in the Palaeozoic rocks, and 

 many of them in the Cambrian. Although 

 these early forms are very different from 

 those which succeeded them in the later 

 geological periods, they do not possess a 

 structure which can be recognized as in any 

 way primitive or ancestral. The Echino- 

 derma are the most distinct and separate 

 of all the Coelomate Phyla, and they were 

 apparently equally distinct and separate at 

 the beginning of the fossiliferous series. 



In concluding this imperfect attempt to 

 deal with a very vast subject in a very 

 short time, I will remind you that we were 

 led to conclude that the evolution of the 

 ancestor of each of the higher animal 

 Phyla, probably occupied a very long 

 period, perhaps as long as that required for 

 the evolution which subsequently occurred 

 within the Phylum. But the consideration 

 of the higher Phyla which occur fossil, ex- 

 cept the Yertebrata, leads to the irresistible 

 conclusion that the whole period in which 

 the fossiliferous rocks were laid down must 

 be multiplied several times for this later 

 history alone. The period thus obtained 

 requires to be again increased, and perhaps 

 doubled, for the earlier history. 



In the preparation of the latter part of 

 this address I have largely consulted Zit- 

 tel's great work. I wish also to express my 

 thanks to my friend. Prof. Lankester, whom 

 I have consulted on many of the details, as 

 well as the general plan which has been 

 adopted. E. B. Poulton. 



OXFOED. 



