PROBLEM OP EVOLUTION — CAULLERY. 333 



Toward 1860 the study of hybridization had led two botanists, the 

 Austrian monk Gregor Mendel and the French botanist Naiidin,^ 

 simultaneously but quite independenth^, to conceptions which did not 

 particularly attract the attention of their contemporaries, but which 

 were brought to light again in 1900, and which then formed the 

 starting point of very many and important investigations. The ex- 

 perimental study of Mendelian heredity has been carried on, espe- 

 cially here in Harvard, with great success by Mr. Castle on various 

 mammals and by Mr. East on plants. This topic, therefore, is famil- 

 iar to the students of biology in this university. I shall speak of it 

 for the present, only to state the general results. Let me recall to your 

 minds as briefly as possible the essentials of Mendelism. According 

 to this doctrine most of the properties which we can distinguish in 

 organisms are transmitted from one generation to another as distinct 

 units. We are led to believe that they exist autonomously in the 

 sexual elements or gametes, and we caft, therefore, by proper crossing, 

 group such and such properties in a single individual, or, on the con- 

 trary, we can separate them. The biologist deals with these unit 

 characteristics as the chemist does with atoms or with lateral chains, 

 in a complex organic compound. The properties which we distin- 

 guish thus are nothing but the very indirect external expression of 

 constituent characteristics of the fundamental living substance of 

 the species. But we imagine, and it is in this that the enormous im- 

 portance of Mendelism consists, that it has been the means of giving 

 us a more precise idea than we have had heretofore of a substantial 

 basis for heredity. In itself Mendelism is only symbolism, like the 

 atomic theory in chemistry, but the case of chemistry shows what 

 can be drawn from a well-conceived symbolism, and the Mendelian 

 symbolism becomes more perfect each day in its form, in its concep- 

 tion, and in its application. The recent works of T. H. Morgan- arj 

 particularly interesting in this respect. 



Further, the facts furnished by Mendelism agree well with those of 

 cytology. The results are explained easily enough, if we accord to 

 the chromatine in the nucleus, and particularly to chromosomes, a 

 special value in heredity. The agi-eemeht of cytology and of Men- 

 delism in incontestably a very convincing fact and a guide in present 

 research. 



But if we rfeturn now to the study of evolution, the data of Mendel- 

 ism embarrass us also very considerably. All that it shows us, in 

 fact, is the conservation of existing properties. Many variations 

 which might have seemed to be new properties are simply traced to 

 previously unobserved combinations of factors already existing. 



1 " Nouvelles Reeherches sur rHybriditc? dans les V6g«aux." Nouvelles Arch, du Mus. 

 Hist. Nat., Paris, Tome 1, 1865. cf. p. 156. 



2Cf. Morgan, Sturtevant, Muller and Bridges, " Tlie Mechanism of Mendelian Heredity ,' 

 New Yorls. 1915. 



