102 FROGHOPPER BLIGHT OF SUGAR-CANE. 



In each case the importance of the dry Indian Summer is emphasised, 

 and in the first two figures there is some evidence of the influence of a 

 longer dry season and wetter June and July in producing blight. 



The beneficial influence of a more regular supply of moisture is well 

 shown by certain observations. Thus in the Princes Town district the 

 rainfall appears to be niore regular, and less subject to droughts than 

 the rest of the sugar-cane areas, and it is this district that suffers least 

 from blight. Another still more striking case is in a field on Hermitage 

 Estate which is regularly damaged b3? blight, but, apparently owing to 

 the gradual seepage, the canes immediately round a pond at the top of a 

 ridge and in a line down the slope below it are always much less 

 damaged than those on either side. 



Similar conclusions liave been arrived at in British Guiana by 

 Bodkin and Moore, both of whom believe that the severe outbreak of 

 blight in that country in 1918 was brought about, or at least influenced 

 by abnormal weather conditions, the most important of which was 

 unusually wet weather at the beginning of the year, which produced a 

 waterlogged condition of the soil and backw'ard cane plants. 



The total effect of the rainfall can be brought about in several 

 different ways, chief of which are: — 



(1.) Direct effect on the froghoppers. 



(2.) Indirect effect on the froghopper through its enemies. 



(3.) Direct effect on the canes. 



(4.) Indirect effect on the canes through root fungi. 



(1.) Direct effect on the froghopper. 



The froghopper cannot continue its cycle without a certain amount 

 of moisture, and except in wooded country (forest and cacao districts), 

 along the banks of streams, and occasionally in field drains, the con- 

 ditions in the normal dry season in Trinidad are below the minimum 

 moisture required for breeding. The insect as a i-esult spends this part 

 of the year chiefly in the egg stage, which has been shown experimentally 

 by Guppy to be capable of resisting at least six months desiccation. 



The first rains of the wet season are the signal for the hatching of 

 the eggs and resumption of activity, and as these occur more or less 

 simultaneously over the island the dates of the broods of the froghoppers 

 bear a constant relation to the first rains. 



In Fig. 25 is tabulated all the information that can be obtained for 

 the fourteen years 1906-1919 with regard to the beginning of the wet 

 season and the occuirence of the broods. In 1908, 1909, 1915, and 1919 

 the wet season started at the end of April and the beginning of May. 

 In all these years (except in 1908, for which we have no records), the 

 first brood had begun to appear before the end of .Tune and the second 

 brood before the end of August. In 1911 and 1913 on the other 

 hand the wet season did not start till the last week in May and in both 

 these years the first brood was at the end of July and lasted into August. 



The beginning of the wet season is sometimes well marked and some- 

 times gradual, the latter usuallj' when it has started early. Allowing 

 as far as possible for this lack of definition it may be stated with 

 some degree of accuracy that the first brood is at its height 53-61 

 (with an average of 57) days after the first rains. 



It is more difficult to estimate the maximum of the second brood, 

 but it seems to occur a few days under two months (about 58 days) after 

 the first, with the third brood at a similar interval later again. 



