OF THE LARVA OF PYG^RA BUCEPHALA. 189 



to give up this task, and do not think that it can be effectually accomplished until we 

 know the arrangement which exists in some of the intermediate genera. 



Some, however, of the more remarkable peculiarities of the thoracic segments in 

 Pygara are equally present in Cossus. Thus, the altered position of 1, 8, 8', 18, 22 and 

 82 resemble those of D, C and E, c and e and C+. The remarkable muscles, however, 

 marked 76 and 77, do not appear to have any close representative in Cossus. 



This comparison, however, is made far more difficult by the arrangement of Lyonet's 

 figures, and by the fact that the same letters are certainly not always used for the corre- 

 sponding muscles in different segments. 



In the first abdominal segment of the larva of Foniia rapi the larger muscles were 

 arranged nearly as in Tyfjccra, except that 2 was attached to the skin at the posterior end 

 of the segment, as in the larva of Cossus. In the larva of Disxjhrugis cceruleocephala the 

 muscles 1 and 2 on the one hand, and 16, 17 and 18 on the other, were represented by 

 numerous separate fascicles, amounting to at least ten in each instance. 



Tills sepai-ation of the fibres composing a muscle into separate fascicles is carried to a 

 much greater extent in the larvae of Coleoptera, or at least in Dyticus and the wood-feed- 

 ing Lamellicorns, which alone I have examined. In these two groups each of the larger 

 muscles is represented by at least twenty separate fascicles, which makes it far more 

 difficult to distinguish the arrangement of the muscles. 



Tlie reserves of fat in the larvae of these Lamellicorns are stored up in large vesicles, as 

 much as -01 of an inch in diameter, and which, being connected together into thin mem- 

 branes, like a mesentery, have a beautiful bead-Kke appearance to the naked eye. 



The muscular system of the larva of Tiptda oleracea and of Ctenophora bimaculata* offers 

 us many interesting points of difference. Beiag unprovided with legs, these little creatm-es 

 move by resting one part of their body against some solid object, and then pushing the 

 anterior part forwards. To enable them to do this with facility, an immense number 

 of muscles are attached to the inner side of the skin. The total number, indeed, faUs 

 a little, though very little, short of that in Cossus ; yet the real complexity is greater, 

 since the average number of muscles in each of the body-segments is rather over seventy 

 in Tipula, while in Cossus it is about sixty. This sm-ely shoAvs greater complexity than 

 the larger total, which is merely made up by the irrelative repetition of a lesser number 

 of muscles in a greater nmnber of segments. 



The joints of the back in the larva of Cossus, like those of Fygcsra, allow much less 

 play than those of the ventral side, and we find the muscles entirely in accordance with 

 this structure. On the contrary, in the larva of Tipula and of Ctenophora bimaculata, 

 the back can be bent almost, if not quite, as much as the belly ; and if this were not abeady 

 knoAvn, it might be deduced from the arrangement of the muscles. 



Lyonet remarks (p. 154, I. c.) that his " o," which corresponds to my 21, "est remar- 

 quable, en ce que, pendant que les autres muscles droits se terminent aux divisions de lem-s 

 anneaux, son extremity postdrieure passe au 4, 5, 6, 7, 8, et 9 anneaux cette division, 

 et s'insere assez avant dans I'anneau qui suit; ce qui vraisemblablement a 6t^ ainsi 



* Mr. Walker was kind euo\igh to name for me specimens bred from larvse resembling, and found with, tliose 

 wbieb I dissected. 



2c2 



