784 



SCIENCE 



[N. S. Vol. XXXIII. No. 855 



should be nothing new; nothing surprising. 

 We know already what to expect in any 

 simple mating of different races of silkworms. 

 As regards larval pattern and cocoon color 

 the inheritance behavior is usually Mendelian. 

 " Moricaud " (all dark) larval pattern is domi- 

 nant over " tiger," or banded, pattern ; banded 

 pattern is dominant over unhanded (all light). 

 Yellow cocoon color is dominant over white. 

 And the relation of dominant and recessive is 

 of the usual Mendelian character in Fi, F,, Fj 

 and succeeding generations. Now although 

 there are two kinds (two races representing 

 these alternative larval and pupal characters) 

 of spermatozoa in the spermatheca of double- 

 mated females, presumably but one spermato- 

 zoid finds its way into the egg, and fuses its 

 nuclear matter with the egg nucleus. That is, 

 the female, although double-mated, is pre- 

 sumably only single-fertilized. 



As a matter of fact the inheritance be- 

 havior of the Fj and succeeding generations 

 derived from these double-mated females does 

 not seem to bear out the simple presumption 

 just stated. The presence in the body (sper- 

 matheca) of the female of two kinds of 

 spermatozoa seems to disturb matters. The 

 comfortable simplicity and regularity of 

 Mendelian inheritance fails to maintain itself. 

 The troubles of irregularity which have not 

 been wholly wanting even in single mating 

 silkworm experiments — and which I have 

 termed in my account of several years' experi- 

 ence of these matings, " strain and individual 

 idiosyncrasies," a term not looked on with 

 favor by thorough-going Mendelians — these 

 irregularities are accented in the double-ma- 

 ting experiments. The irregularities indeed 

 almost assume a seeming of regularity; a non- 

 Mendelian regularity, if such a heresy is ad- 

 missible. 



I shall not try to give here the full data of 

 my 85 "double-mating" lots of 1910. But I 

 shall give a considerable number of examples 

 and a general statement of these results. 

 Later, if worth while, all the data can be 

 given ; together, I may add, with the results of 

 three or four years' more work in single- 

 mating crossings to test further the inherit- 



ance of certain egg, larval and cocoon char- 

 acters, already pretty fairly determined by the 

 original series of several years whose results 

 have been published. 



I shall limit the examples referred to in this 

 paper to matings among three races and shall 

 refer only to cocoon characters. The three 

 races are Istrian Yellow, a strong Austrian 

 race producing large golden yellow cocoons; 

 French Yellow, a race producing smaller, 

 salmon yellow cocoons; and Bagdad White, 

 a Turkish race producing large pure white 

 cocoons. Bagdad White is a race whose white 

 cocoon color, instead of being regularly reces- 

 sive to yellow in crossings with yellow cocoon- 

 ing races, is sometimes dominant, as I have 

 shown in my 1908 report (pp. 24r-25, section on 

 "strain and individual idiosyncrasies"). 

 All these races have been bred as pure races 

 by me for the last ten years; that is, races 

 faithfully transmitting certain larval and 

 cocoon characters. 



In 1908 a Bagdad White female was mated 

 with a French Yellow male from 9 :55 a.m. to 

 11:55 A.M. and with a Bagdad White male 

 from 11:55 a.m. to 1:55 p.m. The young, 

 reared in 1908, were all white cocooners. One 

 mating (1908) (single) among these young 

 produced (1909) 111 white cocoons and 44 yel- 

 low cocoons; another produced all white 

 cocoons; also another produced all white 

 cocoons. A mating (1909) between two white 

 cocooners out of the 111 white and 44 yellow 

 lot, produced (1910) all white cocoons. A 

 mating of two yellow cocooners produced 

 (1910) 8 white cocoons and 40 yellow cocoons; 

 another produced 9 white cocoons and 29 yel- 

 low cocoons. A mating (1908) between two 

 white cocooners of one of the all white cocoon 

 lots produced (1909) all white cocoons; and 

 so did another. A mating (1909) of two white 

 cocooners of one of these all white cocoon lots 

 produced (1910) 15 white cocoons and 2 yel- 

 low cocoons (sick lot) ; another produced all 

 white cocoons. A mating (1909) of two white 

 cocooners from the other all white cocoon lot 

 produced (1910) 17 white cocoons and 4 yel- 

 low cocoons; and another produced 14 white 

 cocoons (both small sick lots). 



