162 PEOr. W. H. JACKSON ON THE 



The development of tlie second portion is very rapid indeed. Tlie liollow base 

 (PL XVII. fig. 50, PI. XVIII. fig. 52, X , X ) upon which the first vesicle is raised extends 

 forwards, but it does so apparently only for a certain distance. At present exact details as 

 to the mode of growth at the apex of this portion are not clear to me. This much is certain 

 that while near the base of the vesicle there is a tube with walls formed of a single 

 layer of cells, the tube as it nears the larval oviducts becomes solid, and is more or less 

 confounded with the hypodermis. My own impression is that there is an ingrowth from 

 thehypodermis of cells which subsequently arrange themselves in the form of a tube ; in 

 other words the invagination of cells is at first solid. This impression is borne out by 

 tiie section shown in fig. 56, PI. XIX., which comes from a quiescent caterpillar. The 

 cells of the hypodermis are at this time considerably elongated, at least in the posterior 

 portion of the body of the animal. Moreover, they lie sloping forwards ol)liquely in the 

 region of the oviduct and are consequently always cut at an angle in a transverse section. 



The posterior portion of the azygos oviduct develops as a widely open furrow 

 extending backwards from the anterior vesicle to the posterior vesicles, with the basal 

 portion of the cavity of which it is continuous ; in fact it is seemingly formed by a 

 continuation of the two folds wliich elevate the anterior vesicle (PI. XVII. fig. 50, 

 PL XVIII. fig. 52, X, x). By the time the caterpillar is ready to pupate, the furrow is 

 converted into a tube by the fusion of its edges. 



The tubular azygos oviduct thus constituted opens to the exterior in two places, by a 

 small aperture beneath the anterior vesicle, which coincides usually more nearly to the 

 base of the recej)taculum seminis than to tlie base of the bursa copulatrix, and by a 

 second aperture just in front of the anterior end of the sebaceous region of the posterior 

 vesicles. A well-marked external or superficial furrow corresponds to nearly the whole 

 extent of the azygos oviduct, even after its complete establishment. The chitinoid 

 cuticle dips into this furrow, which is thus the cause of the longitudinal line crossing 

 the sternal regions of the eighth and ninth somites of the chrysalis. 



The facts stated above are illustrated in the series of figures given in PL XVIII. 

 figs. 54-, 55. 



The larval oviducts are taken up by the anterior extremity of the first section of the 

 azygos oviduct. Their ventral ends are at the close of this period perfectly separate 

 from the hypodermis at the spot where the hypodermic ingrowth of cells to form the 

 azygos oviduct becomes connected to them. It seems to me that they grow slightly at 

 their ventral extremities towards one another ; for the distance between the two ventral 

 ends in the active caterpillar is certainly diminished in the quiescent caterpillar about to 

 pupate. They are still solid in cross section at this time, and possess a single layer of 

 superficial nuclei (PL XIX. fig. 59). A surface view of the ventral ends of the two ducts 

 (PL XIX. fig. 58), taken from a caterpillar which had just entered on the quiescent 

 stage, lends support to the idea that the oviducts grow ventrally. 



If a section of the hypodermis at this stage, made more or less parallel to the 

 surface, is carefully stained, it wiU be seen that it consists of distinctly darker arese 

 surrounded by lighter lines. The darker areoe, I believe, represent the bodies of the 

 hypodermic cells connected by a small amount of intercellular substance (PL XVII. 



