520 



SCIENCE 



[N. S. Vol. XXXVI. No. 929 



This book is one of several from German 

 authors that have been welcomed by American 

 students because of the want of a suitable 

 text in our own language. In twenty chap- 

 ters the author treats of variation and sta- 

 tistical methods, with their present and 

 former application, mutations and fluctua- 

 tions, inheritance of acquired characters, 

 Mendelism, mechanism of heredity, graft 

 hybrids, and sex-determination. 



Statistical methods are discussed briefly, 

 not as in a text-book of statistics and their 

 manipulation, but merely to point out the 

 kind of problem amenable to attack in this 

 way. The author is, however, cautious in ad- 

 vocating statistical methods, pointing out as 

 others, particularly Johannsen, have done, 

 that mathematics without biological analysis 

 is valueless. At the close of the discussion of 

 biometrics, Galton's laws of regression and of 

 ancestral contributions are briefly reviewed, 

 but the view is advocated that these laws are 

 no longer of biological value. In the conclu- 

 sions drawn from the study of pure lines, the 

 author is in substantial accord with Johann- 

 sen. 



Mutations are regarded as probably not es- 

 sentially difllerent from fluctuations, the latter 

 being capable, under certain circumstances, 

 of becoming the former. This conclusion is 

 based largely on the work of Tower. The 

 question of mutations then becomes. What 

 fluctuations may become heritable, and under 

 what circumstances? 



Prefatory to a discussion of acquired char- 

 acters, much evidence is adduced to show a 

 connection between somatic and germ cells. 

 In this evidence is included the effects of 

 gonad extracts upon somatic characters, 

 though one may wonder what bearing this has 

 upon somatic influence on germ cells. One 

 misses the negative results obtained by Smith 

 from injections in fowls, but the book was 

 probably written too early to include this. 

 Guthrie's transplantation experiments are ac- 

 cepted at par, though there seems good rea- 

 son to doubt their validity. 



The inheritance of acquired characters is 



believed, on the strength of the newer evi- 

 dence, to be probable. Goldschmidt is the 

 more ready to adopt this view because he re- 

 gards mutations as only special forms of 

 fluctuation. The theory of parallel induction 

 (the simultaneous action of the environment 

 on both soma and germ cells so as to produce 

 identical variations in parent and offspring), 

 which was supported by Tower's experiments 

 on Leptinotarsa, meets, in the author's opin- 

 ion, an insurmountable difficulty; but how 

 simple induction (the effect of an already 

 modified soma on the germ cells so as to pro- 

 duce the same modification in the offspring) 

 encounters any less insurmountable difficul- 

 ties, is not explained. They seem to the re- 

 viewer to be equally inconceivable. 



The half dozen chapters on Mendelism con- 

 stitute a treatise, for the most part elemen- 

 tary, in which are discussed the various cases 

 of mono-, di- and poly-hybrids. Doubt is ex- 

 pressed as to whether species hybrids exhibit 

 Mendelian behavior, on the ground that 

 equality of reciprocal crosses is a rule of 

 Mendelism, whereas in many species crosses 

 the reciprocal crosses are distinctly unequal. 

 The reviewer finds cases where reciprocal 

 varietal crosses are also unequal. To explain 

 intermediate Fj the author accepts as prob- 

 able the assumption that many genes com- 

 bine to produce a single visible character. 



The Sutton-Boveri hypothesis of the segre- 

 gation of genes is rejected on the ground that 

 there are cases in which there are more inde- 

 pendently heritable characters than there are 

 chromosomes, notwithstanding the fact that 

 the existence of such cases has not been 

 demonstrated to the satisfaction of all genet- 

 icists. In its place the author adopts a theory 

 of end-to-end union of the chromosomes 

 (telosynapsis) in such a way that either 

 maturation division may be a reduction 

 division, whereby the number of possible 

 combinations is greatly increased. The indi- 

 viduality of the chromosomes is not main- 

 tained in this theory. 



The chapters on sex-determination are per- 

 haps one of the least satisfying parts of the 

 book. This is partly owing to the unsettled 



