NOVEMBEB 25, 1910] 



SCIENCE 



IZl 



2 ^ pallida X 2 J decemlineata 



i 

 Hybrids, decemlineata characters dominant 



5(?X 3? 



Pallida 

 6 c?, 4? 



Pallida 

 ISc?, 23? 



Decemlineata and hybrids 

 16 c?, 14? 



Pallida 

 8 c?, 7? 



Decemlineata and hybrids 

 26 c?, 28? 



This is a much more detailed experiment 

 than those of Standfuss, Llerrifleld and 

 Fisher, and it shows that the changes pro- 

 duced by the action of altered conditions 

 on the maturing germ-eells were definite 

 and discontinuous, and therefore of the 

 nature of mutations in De Vries's sense. 



In another experiment Tower reared 

 three generations of decemlineata to test 

 the piirity of his stock. He found that they 

 showed no tendency to produce extreme 

 variations under normal conditions. From 

 this pure stock seven males and seven fe- 

 males were chosen and subjected during 

 the maturation periods of the first two 

 batches of ova to hot and dry conditions. 

 Four hundred and nine eggs were laid, 

 from which sixty-nine adults were reared, 

 constituted as follows: 



Twenty (12 c?, S?) apparently normal decem- 

 lineata. 



Twenty-three (lOj?, 13?) pallida. 



Five {2 c?, 3?) immaciilothorax. 



Sixteen (9(?, 7?) alhida. 



These constituted lot A. 



The same seven pairs of parents sub- 

 jected during the second half of the repro- 

 ductive period to normal conditions gave 

 S-iO eggs, from which were reared 123 

 adults, all deccmlineatas. These consti- 

 tuted lot B. The decemlineatas of lot A 

 and lot B were reared side by side under 

 normal and exactly similar conditions. 



The results were striking. From lot B 

 normal progeny were reared up to the 

 tenth generation, and, as usual in the 

 genus, two generations were produced in 

 each year. The decemlineatas of lot A seg- 

 regated into two lots in the second genera- 

 tion. A^ were normal in all respects, but 

 A^, while retaining the normal appearance 

 of decemlineata, went through five genera- 

 tions in a year, and this for three succes- 

 sive years, thus exhibiting a remarkable 

 ph3'siological modification, and one without 

 parallel in nature, for no species of the 

 genus Leptinotarsa are known which pro- 

 duce more than two generations in the 

 year. This experiment is a sufficient refu- 

 tation of Weismann's argument that the 

 inheritance of induced modifications in 

 Vanessa urticcB is only apparent, the phe- 

 nomena observed being due to the inheri- 

 tance of two kinds of determinants — one 

 from dark-colored forms which are phy- 

 letically the oldest and the other from 

 • more gaily colored forms derived from the 

 darker forms. There is no evidence what- 

 ever that there was ever a species or 

 variety of potato-beetle that produced 

 more than two, or at the most, and then as 

 an exception, three broods in a j^ear. 



The modified albinic forms in this last 

 experiment of Tower's were weakly; they 

 were bred through two or three generations 



