166 TBINIDAD AND TOBAGO BULLETIN. [XVIIL 3. 



DISCUSSION. 

 It will be seen that there is strong evidence that blight is more 

 wide-spread in years or localities with unusual alternations of wet and 

 dry periods and particularly when the Indian Summer dry period in 

 September or October is well uiiirked. 



The relation of this fact to the general problem of understanding 

 and controlling blight will be dealt with more fully later in a report 

 which will take into consideration other associated conditions. The 

 object of the present report is chiefly to give accurate data for future 

 discussion. 



The follow'ing suggestions however, arise from a study of this 

 relation. 



The critical period at which the canes seem to need moisture to 

 prevent blight is in September or October. This is in most years just 

 after the second brood of froglioppcrs. Drought at this time cannot 

 aiTect the number of froglioppers of the earlier broods. It may affect 

 the numbers of the third brood (possibly by preventing the spread of 

 tlie green muscardine fungus) and so increase the damage later in the 

 year. But probably the most important effect is that it prevents the 

 recovery of the canes to wdiich damage has been done by the first two 

 broods, and in those frequent cases where root fungi are also present, 

 it works on an already diseased root system and causes the more or 

 less complete drying up of the canes. 



The fact that this important determining condition occurs after 

 the second brood is further confirmation of the position to which we are 

 driven by other considerations also — namely, tliat the actual number of 

 froglioppers present in afield, although naturally not to be ignored, is 

 often of less practical importance than the general ability of the field 

 of canes to tvitlistand the damage. 



In a previous report on the i-elafcion of Uoot Fungi to Blight 

 {Bull. Dept.Agr. XVIII. 1919. p. 52-6) attention has been drawn to the 

 same point and figures have been given showing the number of frog- 

 hoppers that may be present in a field without causing permanent 

 injury, and also showing that many fields have been destroyed with 

 but a small proportion of this number. 



It may be argued that the rainfall is a condition over which the 

 planter has no control. This is true of the actual rain wliicli reaches 

 the ground, but the jn-oportion of this which becomes available to the 

 \}\a.xxi depends very largely on agricultural operations. 



We have seen that low rainfall, and to a less extent high rainfall, 

 are both conducive to blight. There is one agricultural process which 

 will reduce both evils and that is Drainage. 



Good drainage carries off surplus water in time of floods and also 

 (and this effect is less generally realised) provides more available 

 water in time of drought. Drainage lowers the water level in the soil, 

 and thereby makes a deeper layer of the soil available for the root 

 system of the plant, from which more moisture can be extracted in time 

 of drought. 



i 



