MORPHOLOGY OF THE LEPIDOPTERA. 155 



their point of attacliment between tlie seventh and eightli somites as two separate 

 broader and longitudinally striated l)ands, Avliich end in contact with two small white 

 knobs. They shorten gradually in the quiescent caterpillar, and are thickened at the 

 commencement of pupal life. By their union they form the vagina. The two knobs 

 just mentioned are, at the outset, separated from one another by a slight space, and are 

 fixed to the hypodermis (Sehleimnetz). A nerve derived from the terminal ganglion of 

 the chain passes to each of thiMU. Whilst the formation of the vagina is taking place, 

 the knobs approach one another and unite; moreover they are brought by the contrac- 

 tion of the muscles and skin in pupation into opposition with the attached ends of the 

 two oviducts, a statement which seems somewhat at variance with what is said as to the 

 origin of the vagina. They are destined to give rise to four swellings, the rudiments of 

 the bursa, the receptaculum, and tlie two sebaceous glands. The latter, with a want of 

 consistency, are said to be derived from " the knobs themselves." Suckow, like Herold, 

 does not explain how there come to be two oviducal apertures, though he knew them to 

 be present, uor do his two plates (Taf. v., vi.) afford any clue to a solution of the 

 difficulty. 



. My own investigations now to be detailed have been carried out entirely on Vanessa lo. 

 It was my intention to re-examine Fierls hrassicce, but unfortunately there was a great 

 scarcity of this butterfly at the beginning of the year 1889, and neither my own 

 exertions nor the exertion of friends could procure a sufficient supply. I hope, however, 

 to supplement next year the material now in my possession, and to give the results in 

 my next contribution. Of Vanessa lo a large supply of specimens was obtained on 

 July 10th and 14tli from one and the same locality on the banks of the Isis. The 

 history of these specimens may be shortly stated in the form of a table : — 



Killed as larvte by TachincB 14 



Dissected and preserved as larvse 44 



Killed by accidents in pupation 5 



Pupte of undetermined sex 17 



Male puptv 123 



Female pupai 137 



340 



In addition to the fourteen larvte killed by Tachhice, two pupae subsequently perished 

 by the same means. As to the pupa? of undetermined sex, it may be explained that if a 

 pupa fails to catch hold of its pad of silk and consequently falls to the ground, owing to 

 the strong conti-action of the muscles which sets in soon after pupation, the abdomen 

 becomes curved towards the ventral aspect, and the eighth and ninth sterna are drawn 

 beneath the seventh sternum, so that the sexual apertures are completely hidden. 



There are two points of biological interest connected with the pupse to whicli I should 

 like to draw attention briefly. 



The tirst is the large proportion of females as compared witli males, which contrasts 

 greatly with what was found by Bessels in other species. Even supposing that the 



