270 



ceive only occasional showers. They are also not infrequently 

 found in the forests where the slopes are abrupt and in -ome 

 instances these gi-asses are to be found growing at only a few 

 feet above sea level. All the species when gi-owing in the 

 wind-swept regions have long drooping leaves and rather 

 short stout stems. They are always in bunches or tussocks, 

 often a foot across. 



These grasses appear to have attracted but little attention 

 from the early entomologists. Blackburn found one species 

 of CoryJophus, which he recorded as occurring on grasses 

 which probably occurred in bunch grass. 



Mr. O. H. Swezey first found an endemic insect connected 

 with bunch grass in 1906. which Kirkaldy descril)ed as 

 Kelism swezeyi without indication of the host plant and with- 

 out recognizing its endemic nature. Ten years later a second 

 specie* of KeJusia was foimd by Mr. Swezey in June. 1916, 

 and described as Eelisia em-oloa by Mr. Muir. Since that 

 time much attention has been devoted to collections upon 

 Eragrostis by Swezey, Timberlake, Ftillaway. Giifard and the 

 writer. These investigations have shown that the Inmch gi-asses 

 have a peculiar fauna of their own with several clearly en- 

 demic species, and that the tussocks furnish shelter for several 

 other insects. This fauna has lieen best worked on Oahu but 

 a few species have been studied elsewhere. It is noticeable 

 That many of the species found in bimch grass also affect the 

 sand-binding grass, Sporobolus virginicus. 



Our present knowledge of the entomology of the bunch 

 grasses may be summarized as follows: 



Coi.EOPTEEA. 

 COCCINELLIDAE. 

 Diomus discedens (Sharp). 

 This species was originally found by Blackburn and 

 described by Dr. Shaiii as a Scymnus but indicated as proba- 

 bly fonning another genus. Mr. Timberlake, who has studied 

 the species, permits me to use his unpublished reference of this 



