160 PEOF. W. H. JACKSON ON THE 



it is divided by a furrow into two almost pear-shaped lialves, tlie broader ends of the 

 pears being apposed, their stalks represented by filaments, usually double, rarely single 

 or incompletely double. The two bodies, the anterior in the eighth somite, the posterior 

 in the ninth, are completely independent one of the other; the ninth intersegmental 

 furrow crosses between them. The furrows which divide them into two halves, a right 

 and a left, indicate their bilateral character. In Pieris bmsslcce, which possesses paired 

 bodies of exactly similar character, tlie furrows are deeper, and in Phalera bucephala 

 the two halves are at first independent and separated by a space. They may be 

 originally independent in Vanessa and Pieris, but if so it is at a stage earlier than any I 

 have seen. Transverse sections of the bodies show that they are formed by wide- 

 moutlied and shallow invaginations of the hypodermis. The filaments attached to each 

 half of the posterior vesicle are also attached at their opposite extremities to the rectum. 

 The anterior filament consists of connective tissue, inclosing a trachea. The posterior 

 consists similarly of connective tissue, but it contains a well-defined band which looks 

 like degenerating muscular tissue. In my specimens no cross striae are visible in it, and 

 consequently its nature must remain for the present undetermined. Both filaments 

 spread in a radiating fashion over the roof of the vesicle. They atrophy away and 

 disappear entirely by the time the larva becomes a pupa. 



The paired bodies of this stage are seen on PI. XVI. fig. 29, and on an enlarged scale 

 PL XVII. fig. 33. The series of sections PI. XVII. figs. 50 «, b, c, d, belong to the earliest 

 stage in my possession, PI. XVIII. figs. 52 a, b, c, to a later stage. They are taken from 

 the first pair of vesicles and show very clearly their paired character. In certain of the 

 sections (figs. 50c, 50 rf) a small lateral fold (x) may be noticed. Comparing them with 

 the corresponding sections of an older stage, PL XVIII. figs. 52 a, 52 b, it will be noted 

 first that the lateral folds ( X ) have increased in depth, and tliat the vesicle itself is 

 perched upon them, is deeper and somewhat less evidently paired. The lateral folds in 

 question are the first indications of the median portion of the azygos oviduct ; at present 

 they extend no further than the base of the vesicle. Figs. 51 a, b, PL XVII., represent 

 sections of the posterior pair of vesicles of the early stage, and they differ but little 

 from the corresponding sections (figs. 53 a, b, c, cl, PL XVIII.) taken from the later 

 stage. 



It is genei'ally said that no cell-outlines are distinguishable in the hypodermis of 

 Insecta, and that it consists of a layer of protoplasm with scattered nuclei overlying a 

 homogeneous basement-membrane (see A. Schneider, Zool. Eeitrage, ii. pt. 1, pp. 82, 

 83), but if a reference is made to PL XVI. fig. 32, it will be seen that cell-outlines 

 are very distinctly indicated in the hypodermis of the caterpillar of F. lo. And it seems 

 to me in consequence that the lines crossing the sections of the vesicles at this stage do 

 really indicate cell-outlines. Further particulars as to this point are given in the 

 account of the next or second stage of the development. 



