5i8 



J. K A WAR A. 



worms, and further, the opening- of the sjiinning process on the 

 labium of the former must also be wider than that of the latter. 

 These differences seem to be closely correlated to the length and 

 diameter of the filament of the cocoons of the various races. 



The strength of the filament of a cocoon is proportional to its 

 diameter, but there are exceptions to this rule. Thus by comparing 

 the strength and diameter of the filament of Awobiki, which are 8.1 

 grams and 2g.i n respectively, with those of Koishimaru which are 

 9.2 grams and 25.8 /<., it will be found that notwithstanding the 

 smaller diameter of the latter its strength, which is estimated at 

 9.2 grams, is far superior to that of Awobiki, estimated at 8.1 grams. 

 Filament of smallest strengths are found among the bivoltins. 

 Thus the filament of the second breed of Kumako (bivoltins) which 

 measurs 25.7 ij. in diameter (nearly equal to that of Koishimaru and 

 Matamukashi) has the strength of 6.3 grams only, which is far 

 inferior to that of Koishimaru and Matamukashi whose strengths 

 are respectively 9.2 and 8.2 grams. Filament of the smallest 

 strength and diameter I have observed in the second breed of 

 Natsuko (bivoltins), while filament of the greatest strength and 

 diameter is to be find in Hakuriu, the strength and diameter of the 

 latter being respectively 13.7 grams and 32.2 /i. 



The degree of ductility of a filament has some relation to its 

 diameter, but this relation is not so remarkable as that between the 

 strength and the diameter of a filament. It may, however, be said 

 that the ductility depends upon the strength of a filament. Thus, 

 the diameter of the filament in relation to its ductility and tenacity 

 in our aboriginal races examined, is as follows : — 



Races 



Diameter of filament 



Ductility 



Tenacity 



Awobiki 



29.1 n. 





S.I grams 



Matamukashi 



25.9 .. 



1 1.6 ,, 



8.2 ^ 



Onichijira 



27.2 ,, 





9.8 ,. 



Koishimaru 



25.8 ., 



12.5 



9-2 ,, 



Kimayu 



28.7 ,, 



12.2 



9.1 .. 



THK 



PROPERTIES 



OF RAW 



SILK. 



For the examination of the properties of raw silk I took samples 

 from our principal races. P2ach thread of raw silk is comj^osed of 

 5 or 6 filaments, and reeled with Zaguri (Japanese hand reel). 'I" he 

 results obtaiitccl were as follows : — 



