76 FAUNA HAWAIIENSIS 



the apex, the ventral surface of a part of the process resting within the carina. 

 The seventh segment is entirely concealed, and in strong contrast with the eighth 

 shows little specific modification. It is produced at the apex into two processes, 

 one on either side, which are more or less bent downwards. These processes are 

 longer in some species than others, and are not hairy as is the apical process of 

 the eighth segment. 



The genital armature itself is also remarkably similar, even in the most widely 

 separated species. The stipites are thin, somewhat concave inwardly, and simple, 

 there being no trace of a lacinia. On the apical portion they are always clothed 

 with long hairs. The sagittae do not extend as far as the apex of the stipites, 

 being strongly curved downwards a little before that point is reached, or in one 

 or two species (owing to the length of the stipites) considerably before. 



The sculpture of the various species is remarkable for its feebleness. The 

 puncturation is always fine and shallow, and excepting a group of species, which 

 are relatively of very great size, there is in no case any definite abdominal puncturation. 

 This extraordinary similarity of sculpture adds greatly to the difficulty of differentiating 

 the species. The anterior area of the propodeum is but ill-defined, and its posterior 

 face always bears a longish erect pubescence. 



The habits of the bees of this genus exhibit considerable variety. On the 

 coast and dry mountain slopes they form their burrows in the ground, but in thick 

 forests and wet districts their cells are made in dead wood, generally of standing 

 trees. The larval food contains comparatively little pollen, as these bees have no 

 special poUinigerous apparatus, and in cells that I have opened the larva was floating 

 on the liquid food. Such pollen as is collected is swept towards the mouth by the 

 curved hairs on the anterior tarsi, an d regurgitated with the honey. Some species are 

 quite peculiar to the coast region, others to the forests, and a few to the region above 

 the forest line. A few are found in dense and gloomy woods, which bees would 

 hardly be e.xpected to inhabit, and may be seen flying round some bush, where 

 a little sunlight is able to penetrate. 



In the following descriptions I have used the term 'anterior area' for that part 

 of the propodeum called by Smith ' the enclosed space of the metathorax,' the ' basal 

 area' of many other hymenopterists. In the Hawaiian species it is not enclosed, and 

 should, I think, rather be called the apical than basal area. In any case the term used 

 avoids any possible confusion. All the known species, where possible, have been 

 redescribed, the old descriptions being insufficient for their determination. 



The species of Nesoprosopis form several more or less definite groups. One of 

 these is easily distinguished by the evidently punctate abdomen of both sexes, and 

 the large, robust form, at least of the females. N. fuscipennis is typical of this 

 group, to which also must be referred N. andrenoides, a most remarkable species, 

 resembling at first sight a medium-sized or rather small Andrena. It differs from 



