398 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1915. 



that has played an important part at the higher levels of scientific 

 achievement. 



I seem to have been caught imawares in the act of moralizing. 

 If so, let it charitably be set do^^n as an attempt to soften the hard 

 fact that 30 years after the " Origin of Species " Ave found ourselves 

 growing discontented with the existing methods and results of phy- 

 logenetic inquiry and with current explanations of evolution and 

 adaptation. Almost as if by a preconcerted plan, naturalists began 

 to turn aside from historical problems in order to learn more of 

 organisms as they now are. They began to ask themsehes whether 

 they had not been overemphasizing the problems of evolution at 

 the cost of those presented by life processes everywhere before our 

 eyes to-day. They awoke to the insufficiency of their traditional 

 methods of observation and comparison and they turned more and 

 more to the method by which all the great conquests of physico- 

 chemical science had been achieved, that which undertakes the analy- 

 sis of phenomena by deliberate control of the conditions under whicli 

 they take place — the method of experiment. Its steadily increasing 

 importance is the most salient feature of the new zoology. 



Experimental work in zoology is as old as zoology itself; never- 

 theless, the main movement in this direction belongs to the past two 

 decades. I will make no attempt to trace its development; but let 

 me try to suggest somewhat of its character and consequences by a 

 few outlines of what took place in embryology. 



The development of the egg has always cast a peculiar spell on 

 the scientific imagination. As we follow it hour b}'^ hour in the 

 living object we witness a spectacular exhibition that seems to bring 

 us very close to the secrets of animal life. It awakens an irrepres- 

 sible desire to look below the surface of the phenomena, to penetrate 

 the mystery of development. The singular fact, nevertheless, is 

 that during the phylogenetic period of embryological research this 

 great problem, though always before our eyes, seemed almost to be 

 forgotten in our preoccupation with purely historical questions — 

 such as the origin of vertebrates or of annelids, the homologies of 

 germ layers, gill slits or nephridia, and a hundred others of the same 

 type. Now, these questions are and always will remain of great 

 interest; but embryolog}^, as at last we came to see, is but indirectly 

 connected with historical problems of this type. The embryologist 

 seeks first of all to attain to some understanding of development. 

 It was therefore a notable e\'ent when, in the later eighties, a small 

 group of embryologists headed by Wilhelm Roux turned away from 

 the historical aspects of embryology and addressed themselves to 

 experiments designed solely to throw light upon the mechanism of 

 development. The full significance of this step first came home to 



