December 13, 1907] 



SCIENCE 



825 



The central problem of morphology — the causes 

 of changes in form or at least the determination 

 of the conditions under which changes of form 

 occur — will furnish the main theme of this 

 treatise. 



On reading this, one is forced to ask how 

 any adequate consideration of this problem is 

 possible without reference to experimental 

 embryology and form-regulation. As a matter 

 of fact the book treats almost solely of those 

 fields in which the results of experimentation 

 can not as yet be analyzed nor definite conclu- 

 sions reached concerning the relation between 

 conditions and the complex effect. Conse- 

 quently the consideration of the morphological 

 problem is of necessity very general and in 

 certain respects rather barren of results. 



The subject-matter of the book is treated 

 under six heads or sections, viz. : Experimental 

 Study of Evolution, Experimental Study of 

 Growth, Experimental Studies in Grafting, 

 Experimental Studies of the Influence of the 

 Environment on the Life Cycle, Experimental 

 Study of the Determination of Sex, Experi- 

 mental Study of Secondary Sexual Charac- 

 ters. 



The first of these sections, " Experimental 

 Study of Evolution," comprises about half the 

 book and includes chapters I.-XIV. Briefly 

 stated the principal subjects considered are : 

 the influence of external conditions on animal 

 structvire and the inheritance of changes thus 

 produced, the inheritance of acquired char- 

 acters, hybridization and the behavior of the 

 germ-cells in hybridization, inbreeding and 

 selection. 



Only a few points can be taken up here. 

 As regards the inheritance of acquired or 

 somatic characters Professor Morgan points 

 out that experimental data have not, up to the 

 present, supported the hypothesis and main- 

 tains that until some positive evidence is pre- 

 sented we can not accept it as a well-estab- 

 lished theory. 



In the chapters on hybridization special 

 stress is laid on the importance of Mendel's 

 law and much space is devoted to an account 

 of the experiments which bear upon it. In 

 conclusion the author holds: 



That Mendel's law accounts in many cases for 

 the results and is therefore an invaluable acquisi- 

 tion to our method of interpretation; yet in some 

 other cases it is evident that the inheritance is 

 not strictly Mendelian. Used with discretion the 

 law may still unlock many problems (p. 166). 



At various points, however, notably on pp. 

 77 et seq. and again on p. 169, Morgan ques- 

 tions the so-called purity of the germ cells and 

 points out that in various cases it is possible 

 to bring out certain characters which should. 

 not be present if the germ cells are pure irfe 

 the Mendelian sense. He suggests the hy- 

 pothesis of an alternating dominance and re- 

 cessiveness in the germ cells instead of purity^ 

 as a means of accounting for the results. 



As a matter of fact neither this suggestion 

 nor Mendel's law contains any solution of tlie- 

 problems involved. Both are merely attempts, 

 to state in general terms what takes place in 

 certain cases of hybridization and as such con- 

 stitute only a formulation of the problems. 

 Moreover, the chapters on hybridization show 

 very clearly that work along this line has not 

 yet attained the point where the problems in- 

 volved can be clearly and consistently stated. 

 Even Mendel's law, which is commonly re- 

 garded as the most important generalization 

 attained thus far in this field, applies at best 

 only to certain cases and Morgan disputes 

 the correctness of one of its fundamental as- 

 sumptions, viz., purity of the germ cells. 



In a discussion of the phenomena of ma- 

 turation of the germ cells in relation to 

 Mendel's law attention is called to the arbi- 

 trary character of the assumptions required as 

 regards the distribution of characters, and the 

 fact that the chromosomes have not been dem- 

 onstrated to be the sole bearers of hereditary 

 qualities is emphasized. Incidentally it may 

 be noted that maturation is considered as in- 

 volving in all cases a transverse and a longi- 

 tudinal division of the chromosomes, no men- 

 tion being made of those cases in which two 

 longitudinal divisions are believed to occur, 

 although they constitute an argument in sup- 

 port of the author's position. 



While the Mendelian terminology is freely 

 used, we find no attempt to discuss or analyze 

 the terms employed. Much has been written 



