198 [February, 



too big for this. The explanation may perhaps be found in the cir- 

 cumstance that the pupa of Bombt/lius is, if I may so express it, very 

 hygroscopic — one kept for a few days in a pill-box soon shrunk into 

 dimensions much smaller than the Andrena, but when plentifully 

 supplied with water soon regained its original bulk. Though the fly 

 afterwards perfected its develoj)ment within the pupa case, it failed to 

 escape from it, so that this experiment or the other unnatural con- 

 ditions had weakened its vitality. 



It is to be noted that the insect follows the habit of the aculeate 

 and parasitic Symenoptera and other parasites in retaining the visceral 

 contents until it assumes the perfect state — a habit at which I have 

 always marvelled. The larva is a good deal more tough and strong 

 than it looks, and though when at rest, it is straight, it can and does 

 arch itself in a circle, bringing the extremities of the body into con- 

 tact, contrasting much with the absolutely passive larvae which I have 

 said it in some points resembles. 



The pupa of Bomlylius major at first glance suggests a Hepialus, 

 say lupulinus. But a superficial examination shows the very notable 

 distinction of five curious prongs or spines on the ventral aspect of 

 the head, that the leg- and wing-cases are free from the abdomen, and 

 that the anal armature is not of a Lepidopterous type. Length, 15 

 to 18 mm. Colour, semi-transparent pale ochreous-brown, or in some 

 specimens greenish, with dark, almost black, markings, the markings 

 corresponding to some extent to harder portions of tegument. 



The head is black, its dorsum smooth and corresponding tolerably 

 evidently to the two eyes, anteriorly it is continued into two spines which 

 curve downwards, and are wrinkled at their bases but smooth and po- 

 lished towards their sharp-pointed extremities— about the middle (an tero- 

 posteriorly) and where the lateral joins the ventral aspect are (one on 

 each side) two other horns or spines — these project almost directly 

 forwards, but are curved with the convexity forwards. Close to the 

 base of each is a bifid process, one branch of which is a subsidiary 

 hard sharp black spine, the other is paler, softer looking, and more 

 rounded, with a fringe of very short hairs, looking like a palpus. These 

 two anterior and two lateral spines are about 1 mm. in length, very 

 sharp when view^ed laterally, but are somewhat blunter when viewed 

 from the front. There is a fifth horn or spine situated ventrally at 

 the junction of the head and thorax, this is somewhat shorter than 

 the others, slopes somewhat backwards, and is thin and sharp viewed 

 laterally, but viewed in front is bifid, each lobe being flat and rounded. 

 The space between the bases of thes(^five horus is occupied by a 



