Feebuart 3, 1922] 



SCIENCE 



111 



heterozygous genetic formula because the 

 homozygous associations of either lethal block 

 further development. Such a factorial situa- 

 tion would maintain a state of constant hetero- 

 zygosis, the fixed hybridism of an impure spe- 

 cies. The genetical impurity will be passed 

 from generation to generation and in this 

 respect the hybrid will breed true until the 

 relative positions of the lethals are changed 

 by a crossover, or the genetical constitution in 

 these respects is altered by a mutation. A 

 crossover frees at once recessive characters 

 which were suppressed by lethals in homo- 

 zygous condition and the sudden appearance 

 of such recessives will simulate mutations 

 although in reality they are manifestations of 

 a process of segregation. 



The theory of balanced lethals offers such a 

 satisfactory interpretation of the behavior of 

 certain Drosophila material, behavior similar 

 in nature to that of (Enothera Lamarckiana, 

 that Muller was quick to suggest the applica- 

 tion of his results to Qllnothera problems. It 

 should be noted that De Vries as early as 1911 

 offered a hypothesis essentially similar to the 

 theory of balanced lethals to account for the 

 peculiarities of the double reciprocal crosses 

 between (Enotliera biennis and (Enothera 

 muricata, forms which, on strong evidence 

 from the studies of Renner, we now believe to 

 be impure species. Investigations of my own, 

 published in 1917, on these hybrids and on 

 others support the conclusions that lethals are 

 common in tEnothera material, but I believe 

 that conditions are more complex than indi- 

 cated by the conclusions of De Vries and 

 Renner. De Vries in recent papers has also 

 made free use of lethals in offering hypotheses 

 to cover certain results of his breeding studies 

 with CEnothera. 



Although it is not my purpose to discuss the 

 mutation theory of De Vries it does seem 

 important to examine critically the position of 

 this theory as it is affected by the evidence for 

 the existence of impure species that are held 

 to a behavior of pure breeding or almost pure 

 breeding by lethals which suppress the appear- 

 ance of segregates. Lethals are not rare in 

 Drosophila and CEnothera material. There is 

 reason to suspect that they are eonunon mani- 



festations of irregularities in the mechanism 

 of the organism of so serious a nature that they 

 interfere with vital processes at some point 

 in the life history, finally bringing the machine 

 to a standstill with death as a result. The 

 workers with Drosophila seem inclined to be- 

 lieve that much of the phenomena simulating 

 mutation in their material is in reality the 

 appearance of characters set free by the 

 breaking of lethal adjustments which held the 

 characters latent. Well known workers have 

 arrived at similar conclusions for Qllnothera 

 material and are not content to accept as evi- 

 dence of mutations the behavior of Lamarcki- 

 ana and some other forms when they throw 

 their marked variants. 



An entirely new conception of mutation 

 phenomena has grown up with meaning very 

 different from that of the past. Qilnothera 

 material selected by De Vries on the assump- 

 tion that it illustrated mutation in a pure spe- 

 cies proves to be highly impure and in genet- 

 ical constitution exceedingly complex. Prog- 

 ress in the study of mutations must follow the 

 usual course in genetical research and rest 

 upon intensive studies of particular characters, 

 analj^zed and traced through experimental cul- 

 tures and tests of cross breeding, with the 

 assistance of cytology at critical points in the 

 life history, and with constant attention to 

 phenomena of infertility and sterility. From 

 the later writings of De Vries it would seem 

 that the master recognizes the newer trend. 

 Logically mutations appear to be more likely 

 from hybrid stock than from pure lines since 

 heterogeneity of germinal constitution obvi- 

 ously invites chemical and physical modifica- 

 tions that might lead to the origin of new 

 genes or to such changes in old genes as would 

 result in different expressions of former char- 

 acteristics. Of particular import is the expec- 

 tation that lethals most frequently owe their 

 presence to heterozygous conditions since the 

 mixing of diverse germ-plasms seems likely to 

 lead to the breaking down of delicate and 

 vital adjustments in proportions relative to 

 the degrees of protoplasmic confusion, and 

 this means chemical and physical disturbance. 

 The intensive study of specific mutations with 

 its effort at analysis to the last degi'ee is a 



