Silkworm from Japan. 365 
prolegs are light-brown, claspers deep green, margined with rose- 
brown; the two triangular brown patches at the anus are margined 
with yellow and white, numerous short yellow hairs stud the 
surface, especially the anterior and lateral aspects; the spines 
curl from behind forwards. Between the third and fourth moults, 
that is, in the fourth stage, my larvae, to my sorrow, all died, so 
that my descriptions are from this point copied from M. Per- 
sonnat’s work, ‘Le Ver a Soie du Chéne, Bombyx Yama-Mai.” 
The interval between the third and fourth change lasts for seven- 
teen days, four of which are comprised in the last moult, which is 
the most trying and difficult to the larvae. It remains a long time 
after moulting without feeding and even without moving its 
place, often from twelve to twenty-five hours. Under the influence 
of air and food the skin, at first pale, assumes the same tints and 
exhibits the same differences of colour between the French and 
Japanese reared specimens as in the preceding stages. In the 
latter the segments from the fourth to the eleventh are traversed 
by a yellowish-brown stripe, which enters the great brown tri- 
angular patch separating on each side the three portions of the 
anal segment. The head and margin of the claspers are of a 
reddish-brown, tipped with green, and on each side on the third 
segment, and often on the sixth below the spiracles, is placed a 
lovely silvery spot of metallic lustre. In the French specimens the 
lateral stripe is white, but interrupted and hardly visible at the 
commencement of each segment, and changing into green at the 
termination; the head is entirely green, like the rest of the body, 
there are no metallic spots, the lateral tubercles are of a deep 
blue, and spots of the same colour changing into dark green fre- 
quently appear at the base of the claspers; these differences are 
common to the fourth as well as the fifth stage. The worm 
now grows rapidly, becomes enormous, attains a length of 33 to 4 
inches, with a proportionate thickness. The tubercles have all 
disappeared in a turgescence of the skin, which seems thickened 
and appears to imbibe the air through an infinity of pores. The 
metallic spots, which in the Japanese specimens to the number 
of three, four or seven, are seen on each side, one on each seg- 
ment,commencing at the fifth, are in the French specimens entirely 
or almost entirely suppressed. The larva having eaten largely 
during sixteen or eighteen days, according to the temperature, 
becomes transparent, of a pale almost yellow green, as the time of 
spinning approaches. He is now slightly erratic, as if seeking a 
site for his cocoon; having made a choice, he empties his intestinal 
canal, similarly to the Cynthia larva, emitting a large drop or two 
