GAMETIC COUPLING AND REPULSION IN THE SIKWORM. 
117 
I wish to express mv hearty thanks to Mr. Y. Takahashi who has 
given me many valuable suggestions during the study, and has kindly looked 
through the manuscripts. 
I am also indebted to Mr. K. Suzuki who reared and crossed in 1911 
with unusual skill and care various strains of my silkworms in the College 
silkworm-nursery. 
II. Description of the Races. 
The silkworms used in the present experiments are of two races: 
1) Japanese tetra-moulters. 11 
2) Chinese tri-moulters. 15 
1) Japanese tetra-moulters (Japanese normal white). 
Of numerous tetra-moulting races, only two, viz. Aojilcu and the 
“Brown ant” were employed. The former is one of the commonest breeds 
in Japan, amd the latter was one which appeared as a sport of an another 
well-known breed, MatamuJcashi. The characteristic of the latter is that the 
newly hatched larva or “ant” is reddish brown in colour, but not black as 
in the normal strains. Both of the races which I made of use are univoltine, 
their larvae being normal-patterned and spinning pure-white cocoons. They 
have been pure-bred for generations in our nursery and proved to be homo- 
zygous for their larval marking and cocoon colour. 
2) Chinese tri-moulters. 
Iu 1910 we procured an egg-carton from our Chinese friend Mr. NÜ 
in Shan-tung. The population reared from this material proved to be a 
mixture of various strains, not only as to the larval markings but also in 
respect of the cocoon colours and moulting-frequency, and from this mixture 
I have isolated the following strains' 0 in the same year. 
1) The term “tetra-moulters” means those races which pass through four moults, 
while the “tri-moulters” those which undergo only three ecdvses before they spin cocoons. 
2) As to the detailed statements of the larval markings and cocoon colours of these 
races readers are referred to pp. 95 — 98 of this Volume. 
