300 Mr. K. J. :Morton on 



olive-bvowii, eacli segment with a lateral salmon-tinteJ sub- 

 cresceiitic portion. 



B.-T. Avritcs, June 10th : " This species is beconiing 

 common. More restless than the other brown dragon-fly ami 

 does not take possession of a definite perch. Some males are 

 browner red than the typical glowing red, which is most 

 beautiful. Rather w;uy and flits from grass to grass by the 

 side of water." The brown dragon-fly alluded to appears to 

 refer to a condition of Trithemis annu/aia. 



Crocolhemi's servUia, Brull6. 



24 (^ (^ , 11 ? ? , 7. v., 14-17. v., 25. v., 1-10. vi. (Basra). 



The shoulder-stripe has only become faint in a few of the 

 most mature exam[)les. 



In attempting to arrive at a more satisfactory understanding 

 regarding tlie distinctive characters of these species of Croco- 

 t/iemis, a partial study of the genitalia of the second abdominal 

 segment of tlie male has been made by removing the parts 

 from a number of specimens and mounting them in balsam. 

 Tiie results may be briefly summed up as follows : — 



(1) The preparations confirm the accepted view that the 

 outer branch of the hamule is more pointed in C. en/thrwa 

 than in C. serviVui, in which it is more truncate. Those flat 

 ])re|)arations, however, do not give an altogether satisfactory 

 idea of the form of the branch, the jjosition being not quite a 

 natural one. The shape is better understood when the 

 hamule is viewed from the side. 



(2) The apex of the inner branch of the hamule appears to 

 be different in the two species. In the Madagascar prepara- 

 tion of C. erythrcea (PI. XIV. fig. 3) the extreme apex is seen 

 to be slightly notched with a strong subapical tooth. After 

 examination of a number of dried S[)ecimous of C. en/(//r<P(i 

 from widely separated localities, I can say positively that in 

 these the toothed or bifid condition is invariably present, and 

 it is interesting to mention that an cxamj)le from Cherra- 

 punji in my collection included by Ris (Coll. Selys, p. 540) 

 under C. servilia as transitional towards C. erythvcva is, 

 in respect of the hamule, true C. eryllircca. The evidence 

 regarding C. servi/ta should perhaps be stated in a more 

 negative form. In none of the examples of C. servilia 

 examined have T been able to confirm the existence of any 

 tooth, the apex apparently being always simple. These 

 critical characters are, however, somewhat elusive, and unless 

 the hamule is in exactly the right ]-)ositioji the tooth niJiy be 

 overlookcil. In the preparation from which fig. 1 (PI. XIV.) 



