Royal Physical So defy. 43 



(whether male or female) are short like the female, and con- 

 sist only of nine joints. They are perfect miniature repre- 

 sentations of the full-grown female antennae. After the first 

 moult, however, a change was perceptible on the antennse of 

 the specimen bred ; and that change can still be distinctly 

 seen in the cast skin, which has been kept. The third joint 

 has grown longer. No trace of division can yet be seen in it, 

 but if we had the next skin (which we unfortunately want), I 

 feel certain we should find traces of divisions in the joints. 

 We have, however, the third and last skin which was moulted, 

 and we see in it that a great change has taken place. The an- 

 tennae still bear the general short, thick, female form. They 

 are only a little longer; but on counting the joints, we find 

 that there are now eleven joints beyond the third, where before 

 there were only six, as if each had been split in two except the 

 terminal joint, and we find the third joint to contain within 

 itself eleven more new ones. It has become elongated, and a 

 series of striae (they can scarcely yet be called joints) run 

 across it, well defined on the interior side, but not so well de- 

 fined on the exterior margin. These with the two basal joints, 

 on which no change has taken place, make up the twenty-four 

 joints of the male antennas ; and on the insect emerging from 

 the last skin, they rapidly extend themselves to the full-grown 

 size. I may observe, that the multiplication of parts by sub- 

 division (although a mode often adopted by nature in other 

 classes of living animals) is not the usual course followed in 

 the case of insects ; for instance, the lulus terrestris (which 

 may be taken as a type of the Myriapods) has, when it 

 emerges from the egg, only eight segments. These are mul- 

 tiplied by the growth of new segments — six at a time ; but 

 the new segments are not formed by a division of the old, but 

 by generation from the penultimate segment at the terminal 

 space. The interesting skins preserved, in this instance, leave 

 no doubt as to the means by which the segments of the an- 

 tennae have been increased. 



As I have already mentioned, however, we have not the se- 

 cond skin ; the insect ate it up before it could be secured. I 

 am not aware whether this singular act of cannibalism has 



