THE BUD-ROT OF PALMS IN INDIA. 



24] 



leaves were much distorted. The lateral position may have been 

 due to the new shoot pushing out before the older dead leaves had 

 been cast, and being forced by these out of the direct line of growth. 

 In other cases the older leaves of one side only were killed, those of 

 the opposite side remaining healthy. The central shoot had broken 

 ofi and had been replaced by a new shoot of small crumpled green 

 leaves. Nearly all the instances so far noticed in which a new shoot 

 has developed to replace that destroyed by the disease, have suc- 

 cumbed to a second attack in a comparatively brief period. In no 

 case as yet observed has the new growth progressed sufficiently 

 to bear fruit and in only one to form a croAvn of ordinary size. 

 This is one of the indications that the parasite is capable of passing 

 into a dormant condition and renewing its growth at a later 

 period. 



V. — Seasonal prevalence. 



Much attention has been given to an attempt to ascertain the 

 seasonal prevalence of the disease. The cultivators and tappers 

 in several places declare that the deaths are most numerous from 

 December to February, and least in the hot months, March to June. 

 A study of the figures from the villages of Amalapur Taluk in which 

 there are records of the number of new attacks observed periodically 

 does not bear out the first part of this statement. It leaves no doubt, 

 however, that the deaths are least numerous between February and 

 July. 



The following table shows the rainfall at Rajahmundry at the 

 head of the Delta and the relative liumidity at 8 a.m. at Masuli- 

 patam on the sea at the southern limit of the affected area : — 



