clxxvi FAUNA HAWAIIENSIS 



numbers together, but more usually few or scattered specimens are noticed. On one 

 occasion in a sheltered corner of a deep gulch on Molokai scores of examples could be 

 seen at once resting with expanded wings on the dried stems of some low plant, nearly 

 every stem having an occupant. Each day for a considerable period they were 

 observed congregated in this favoured spot, but only during the late hours of afternoon 

 sunshine. This dragon-fly exhibits much variation in colour, size, and other respects, 

 and both general variability and local are evident. 



The twenty-six species assigned to Agrion form by far the most remarkable 

 portion of the endemic dragon-fly fauna. The variability of many of the species, in 

 characters that are considered important, is unusually great. In some species the 

 post-ocular spots may be conspicuous, reduced, or entirely absent, according to the 

 individual. Other varieties in colouration may be still more striking. A. anianrodytH))i, 

 for instance, may be an entirely blackish insect, with much bluish white pruinosity on 

 the head, thorax and the base of the abdomen, or it may be of a more metallic black 

 with conspicuous thoracic pale markings, or in addition to much pale ornamentation 

 the abdomen may be largely red. There is very great variability in size of the 

 individuals in many species. There is much general variation in examples taken in 

 the same locality ; in some cases there is sufficient difference between the aggregate 

 of individuals of a species from one island and those inhabiting another to form more 

 or less distinct races or subspecies. Thus the two species, A. amatirodytum and 

 A. calliphya, are of much smaller average size on Hawaii than on the intermediate 

 islands of the group, and their neuration is consequently affected. 



The neuration of many of these Agrions shows much variation in detail, and this 

 not merely in small points, but the more important features are in a plastic state. 

 McLachlan, when he received material from Blackburn for description, some thirty 

 years ago, formed a new genus for the reception of two species sent to him, calling this 

 genus Megalagrion. This genus was essentially based on the neuration of the post- 

 costal area of the wings, a number of the cellules in this area being doubled, instead 

 of forming a single row. When, however, a more extensive collection of Hawaiian 

 species is examined, it is found that this duplication of cellules exists in all stages of 

 development. Species that normally have a single post-costal row of cells sometimes 

 have a number of these divided to form a double row. In some cases the post-costal 

 area is simple on the wings of one side of the insect, partially double on the other. 

 A similar phenomenon is also shown in the series of post-pterostigmatic cellules, where 

 a similar complication of neuration takes place, some species always having more or 

 less a double row, others having sometimes a single and sometimes a partially double 

 row, or the wings on either side may be different. 



It is clear that increase of size of the insect favours duplication of the cellules, 

 though it does not always necessitate it. Thus the three large species, which would 

 more properly form McLachlan's genus MegalagiHon (if that be retained) always have 



