April 15, 1910] 



SCIENCE 



587 



McCraeken extending over a period of five 

 years seem to indicate that the character does 

 not follow out the Mendelian ratio in hybridiz- 

 ing, but rather that the bivoltine character 

 shows an increasing prepotence over the uni- 

 voltine character in consecutive generations. 

 Miss McCracken interprets this as a reversion 

 to an ancestral condition. 



One of the most significant results obtained 

 by Tower in his work with LepUnotarsa,^ con- 

 sisted in so altering the nature of the germ 

 plasm of his developing beetles by certain 

 " stimuli " that a normally two-brooded form 

 became five-brooded — a condition that was 

 perpetuated in succeeding generations. 



Somewhat along the same line, certain ex- 

 periments carried out in Japan and recorded 

 in an obscure journal" would seem to deserve 

 recognition, if only because of their interest 

 in connection with the above-mentioned work 

 of Tower and Miss McCracken. As it is un- 

 likely that this paper is either accessible or 

 intelligible to the majority of occidental biol- 

 ogists, it may be worth while to give a brief 

 abstract of it, in the hope that some one may 

 be induced to repeat the rather uncritical ex- 

 periments of the Japanese and thereby throw 

 more light on the interesting phenomena of 

 alteration of brood habit. 



The article is entitled " The Artificial Pro- 

 duction of Trivoltine Silkworms from Bivol- 

 tine," and the writer, Mr. E. Tsukai, begins 

 by relating how an experienced silkworm 

 grower named Matsumoto, living in a tovm of 

 Shizuoka-ken, called Uragawa, brought some 

 bivoltine silkworm eggs of a dealer some 

 twenty miles to the north, intending to keep 

 them over the winter and rear them the fol- 

 lowing spring. To his astonishment, after a 

 few weeks, the eggs began to hatch. He 

 thought at first that he had been tricked in his 

 purchase, but on recollecting that the climate 

 of his own town and that from which his eggs 

 had come is quite different, he resolved to sus- 

 pend judgment pending investigation. He 

 found, indeed, that there was a difference of 



' Carnegie Pub., No. 48, 1906, p. 289. 

 'Dai Nihon San Shu Kwai Ho (Kept, of the 

 Sericultural Assoc, of Japan), No. 171, p. 5, 1906. 



five or six days in the hatching interval in the 

 two places, the worms which issue in twelve 

 to thirteen days ordinarily, requiring there 

 some eighteen days. Conceiving that a sudden 

 temperature change might have occasioned 

 this alteration in the physiological habit of his 

 silkworms, he decided to experiment. 



Near Takizawa is a famous cavern, the tem- 

 perature of which varies little from 60° [Fah- 

 renheit?] the year round. Within this cave 

 he placed some eggs of the first brood of a 

 bivoltine race, intending to delay their hatch- 

 ing until the eighteenth day. Eggs placed in 

 the cave three days after laying and kept there 

 nine days, on being removed, hatched out three 

 days later, apparently unaltered by their stay 

 in the cave. Next year (1903) he took up the 

 matter again. Some bivoltine eggs were 

 divided into two lots. The first were " brushed 

 down" (first instar), March 31, pupated May 

 17 and emerged June 7. The eggs of these 

 moths after a two-day interval were placed in 

 the cave (March 24). After 13 days, i. e., June 

 6, they were taken out. Six days later (June 

 12) they hatched. In rearing them, it was 

 found that the cocoons were inferior to those 

 of the second brood. The average cocoons of 

 the second brood run about 270 to the sho 

 (1.8 liters). These rap. about 308. Of these 

 only four or five revealed the trivoltine char- 

 acter. 



The first brood of the second lot were 

 " brushed down " April 15, pupated May 17 

 and emerged June 7. The eggs from these 

 after a two-day interval were placed in the 

 cave. After a stay of nineteen days (June 

 27) they began to hatch in the cave. The 

 worms were " brushed dawn " and reared, but 

 were very thin and " thread-like." Larv£e in 

 the second moult average about 15.4 g. in 

 weight. These did not exceed 10.5 g. They 

 pupated July 23. The cocoons were very light 

 and small, 358 of them bulking the same as 

 255 of the ordinary second brood. 



Nevertheless, these all hatched out as trivol- 

 tine moths. Thus the experimenter's aim had 

 been accomplished. 



It is to be regretted that the Japanese writer 

 does not give more explicit information as to 



