June, 1897 1 Packard: Transformations of Hymenoptera. 81 



abdominal segments which are nearly smooth and very round. The 

 specimen described was not fully grown and was found by Mr. J. H. 

 Emerton, August 13, with eggs of the second brood. 



When the larva has voided all its excrement the tubercles over the 

 whole body become very prominent, extending from low down on the 

 side of the body, forming high, regular, very prominent transverse 

 ridges, which beneath the abdomen are more prominent than on the in- 

 side of the thorax. Length, .40 inch. 



In examining the larvae of H. parallelus and H. iigatus the head 

 only differs, so far as one can tell, by the sides of one species bulging 

 out; in the mandibles of H. Iigatus being longer and slenderer, and 

 the notch below being longer and ending in a distinct seta. The head 

 in the two species is of about the same size ; the clypeus is of the same 

 shape, the head above being a little more divided in H. parallelus than 

 in H. Iigatus. The entire larva of H. Iigatus is much longer and 

 slenderer than that of H. parallelus, and the thickened tuberculous 

 portion of the segments inclined to be a little more prominent. These 

 differences are sufficient to produce changes in form, rendering the 

 identification of the larva easy, but the best specific characters are the 

 differences in size and slenderness of form. The larvae being just in 

 the period approaching the serai-pupa stage, the head is protruded and 

 the segments more or less elongated, as the parts of the pupa growing 

 beneath press out the larval skin in various directions. The ovipositor 

 can not be detected beneath the thin larval skin. 



This larva (the following description applies to it when in the early 

 semi-pupa stage, and there are no hairs yet developed) differs from that 

 oi Andrenavicinamhtmg longer and slenderer in proportion. The 

 antennae are shorter, stouter and more clavate. The mandibles in this 

 stage are not corneous. The maxillae are shorter, the lingua much 

 longer than the tips of both pairs of palpi, which are of the same length 

 as in Andretia. The two tubercles behind the ocelli are unusually 

 prominent. Of the three ocelli, which are arranged at points in an 

 equilateral triangle, afterwards becoming a very slight curved line, the 

 middle one in front is not raised. 



In front of the ocelli, arranged transversely in a slight curved line, 

 are four low, flat tubercles which resemble the ocelli ; these disappear 

 when the pupa becomes mature. The head and front, including the 

 clypeus and labrum, are as in A?tdrena, but the supra-clypeal region in 

 the specimen before me is better marked. The legs are a very little 

 slenderer, and the hind tarsi do not reach nearly to the tip, but only 

 half way, as the abdomen is much less elongated than in Andrena. 



