YOSHIMARO TANAKA 239 
side was normal and female, and the left side plain and male. 
Recently Takahashi (1914) has described four mosaic specimens of the 
silkworm. One was a combination of moricaud right and normal left, but 
was male on both sides. The second was normal-patterned, but had oily skin 
on the left side, and opaque white on the right; it was a female. The third 
specimen showed oily patches on the right side, and was female on both sides. 
The fourth individual was normal in marking and showed no abnormality on 
the dorsal side. On the ventral, however, the left was transparent-skinned, 
the right being ordinary opaque. This individual was a gynandromorph, male 
on the right and female on the left. 
The views of previous authors on the cause of gynandromorph ism are 
pretty divergent. Koveki (191 5) attempts to explain it by partial merogony, 
Lang (1912) by mutation in the sex-chromosome. Goldschmidt (191 2) by 
"Potentialdifferenz" of sex-factors, and MORGAN (1913) by polysperm fertili- 
zation. Having worked recently on a gynandromorph of the fruit-fly, 
Morgan (1914h, 1916) comes to view that "gynandromorphs and mosaics may 
arise through a mitotic dislocation of the sex-chromosomes." 
The silkworm mosaics and gynandromorphs must, I believe, be explained 
by mutations taking place in the course of ontogeny. By the term "mutation" 
I do not like to mean a sudden elimination or addition of certain factor or 
factors, but some reorganization or disturbance taking place among somatic 
cells or chromosomes, by which certain factor or factors are suppressed, or 
suppressed factors called into activity. 
The mutation may affect either non-sexual characters only(simple mosaic), 
or the sex- factor only (simple gynandromorph 0 ), or both (mosaic gynandro- 
morph). 
We have sharply marked off "right-left mosaics" e) on one hand, and more 
or less irregular mosaics 3j on the other. It would not be unreasonable to 
1) Gynandromorphs which are not mosaic in somatic characters at all, are at least theoretically 
possible in the silkworm as in other animals. That such examples have not been described until 
now is perhaps due to the circumstance that they would very likely be overlooked, presenting, as they 
would, to the observer no marked external difference from the normal unisexual larvae. 
2) Mosaics 1, 4, 6, lo etc. 
3) Mosaics 2, 3, 5, 7, 8, g, i 1 etc. 
