BACTEKIAL DISEASES. 19 



Because of this lack of (liscriniiiiation in settling, it will bo readily 

 seen, death duo to overcrowding is not, strictly speaking, always the 

 result of overtleposition, but frecpiently results from the overlap})ing 

 of larvse and \n\pve durnig growth on leaves only moderately infested. 

 Since, after settling, the immature stages do not change their loca- 

 tion, specimens havmg ample room during the early larval stages 

 become so large in the pupal stage, if not before, that they may over- 

 lap each other at the molting period, with disastrous results to the 

 indivitlual beneath. Partial overlappmg of the posterior portion of a 

 pupa docs jiot always result in its death, but death invariably follows 

 the overlapping of the anterior or head end of the body. 



EFFECT OF CURLING AND DROPPING OF LEAVES FROM DROUGHT. 



Data collected duruig an unusual period of drought extendmg 

 throughout the fall and winter of 1906-7 show that curling of leaves 

 as an effect of drought has little effect on the vitahty of the fly at 

 this season. In starch, 1907, pupae of the citrus white fly were 

 observed on leaves which had been curled and dry from the elfects 

 of droughts for more than three months. The leaves were so dry 

 that they felt and tore much like paper, but thej^ soon regained their 

 normal texture after the beginning of the rains. The emergence of 

 the adults on trees affected as here described was delayed for several 

 weeks as compared with unaffected trees, but aside from this there 

 was no apparent effect on the insects. 



Although the curUng of the leaves of citrus trees as a result of 

 drought has not, so far as observed, resulted in checking the white 

 flies, the dropping of the leaves may be decidedly effective in this 

 respect. When citrus trees suffer from the effects of drought to the 

 extent of shedding a considerable part of their foliage, the resulting 

 reduction in the numbers of white flies rarely proves of sufTicient 

 advantage to offset the injury to the trees, and the insects as a rule 

 resume their normal status fully as rapidly as the trees recover. 



BACTERIAL DISEASES. 



While no bacterial disease has been recognized as such in produc- 

 ing the very high rate of mortahty often occurring among the larvae 

 and pupae of both species of white flies, there are indications that 

 bacteria play a more important r61e in this connection than has been 

 suspected, and are at times more beneficial in holdmg the fly in 

 check than are the fungi. The fluctuating effectiveness of the 

 unexplained mortality heretofore discussed, without the visible 

 appearance of any fungous parasite which might be responsible, 

 seems to indicate that some parasitic organism is directly concerned. 



