30 BULLETIN 66, UNITED STATES NATIONAL MUSEUM. 



Two focal points were found around which the highest percentages 

 of parasitism were noted; beyond this no more has yet been worked 

 out. 



The dominant flower is Sitilias caroliniana; it is everywhere. It 

 opens about 7 a. m., and closes shortly after 9 a. m.; hence the col- 

 lecting was all done between these hours. The flowers open in all 

 kinds of weather and the bees visit them notwithstanding; on clear, 

 sunny days they are very alert and fly easily, being disturbed by the 

 slightest shadow, while on cloudy days they may be approached easily, 

 and on damp, misty, or rainy days they rise with difficulty and fly 

 slowly. It was always noticed that parasitized bees, especially those 

 which had contained males, did not start easily, and were of a heavier 

 flight, as though burdened and unbalanced. 



.;. Acceleration or retardation of development. 



Whether the presence of the parasite actually influences the time 

 of development of the host is as yet a matter of some conjecture. 

 Saunders (1850), with regard to Prosopis rubicola bred in 1847, records 

 "those bees which produced parasites being always observed in the 

 imago state before others not parasitically affected," and in an un- 

 published notation opposite this remark Saunders has added: "In- 

 stance of this precocity exhibited also in the first reared in 1851. Ex- 

 ceptional, however." 



The writer has observed, as recorded under the examinations of 

 parasitized Polistes annularis, that the immature wasp larvae which 

 remained longest in the nest were unparasitized. These young were 

 probably immune because no female parasite taken showed a suffi- 

 cient development of eggs to have produced triungulinids in time to 

 attack them. 



In the case of Andrena crawfordi the evidence points decidedly 

 toward acceleration. The first flowers of Sitilias caroliniana appeared 

 about April 24; Serinia oppositifolia had been in flower a week longer. 

 On April 26 the first Andrena were taken. On this and the two fol- 

 lowing days 74 bees were collected and 59 or 79 per cent were infested; 

 68 per cent of the parasites were males, and in only three cases were 

 these still in the pupa state. After April 28 the per cent of parasitism 

 rapidly decreased. » 



4. Alteration of general features. 



Perez has contributed several very valuable articles on the effects 

 of parasitism in Andrena and the author has established the ground- 

 work of the following paragraphs on his work (Perez, 1886). A few 

 observations by Mr. Crawford have been kept separate and inserted 

 after section 5 (p. 34). 



a. Abdomen. — In Andrena trimmerana the abdomen becomes sen- 

 sibly shortened and inflated, and more or less globulose. The atten- 



