764 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. XXV. 



As to cultivation, the cost would be about the same as for a cocoanut 

 tope. Whether watering is really necessary or not Mr. Shubrick does not 

 know but opines that they might do without it. But vide instructions for 

 cultivation printed at p. 688, Vol. XX of the Journal. 



W. B. BANNERMAN, Surgn.-Genl., i.m.s. 

 Madras, 12t?i January 1918. 



[ The Revd. El. Blatter, S.J., has kindly furnished the following- 

 note on the above : — 



'' Experience has shown that, on an average, the rainfall during the 

 flowering and fruiting period should not exceed 5 inches. This condition 

 is fulfilled in the case of Trichinopoly. Its mean annual rainfall is 37 

 inches. Up to the latter part of April, but little rain falls. At the end 

 of this month and in May thunderstorms are frequent, and the average 

 rainfall of May is nearly 4 inches. June and July again are dry months, 

 there being on an average only 6 rainy days in the two months. Rain 

 becomes more freqvient in Angust and September, and reaches its maxi- 

 mum in October, in which month there are about 11 rainy days. Novem- 

 ber and December are also more or less rainy (ex Blanford). As to 

 temperature and humidity, the conditions of Trichinopoly are the best 

 possible. The mean temperature is 82*^ F., and the mean humidity is 63 

 per cent., and in April, the driest month, only 54 per cent. Even in June 

 and Jidy it is only 57 per cent., but in November it rises to 76 per cent. 



It is interesting to note that the trees mentioned by Surgeon-General 

 Bannerman flowered on February and ripened their fruit in June, viz., a full 

 month (or perhaps even more) before the usual flowering and fruiting season 

 in the well-known date-producing countries. A month later would be 

 fatal to flower and fruit 



There is apparently no reason why the cultivation of the date-palm 

 could not be conducted successfully in many parts of the Deccan and 

 Southern India, provided always that there is a dry period sufficiently 

 long for the reproductive action of the palm. Prof. J. N. Eraser informs 

 me that excellent dates have for a long time been grown at Shorapur 

 (Gulberga Dist. of Hyderabad)." ] 



No. XXIV.— FALL OF SEED IN A RAINSTORM. 



Under the above heading in Note XX of the Journal of 15th September 

 1917, Mr. Brook-Fox records a "fall" of grains of a red variety of Jowari 

 during a rain-storm in the Junagadh State, on 12th April 1917. He con- 

 siders that the grains must have been ^' blown in'' ^ fiom a neighbouring 

 province. 



I venture to question the probability of this explanation and to ofl'er 

 another. The dispersal of seeds by wind is, of course, a recognized scien- 

 tific fact. But the phenomenon is confined either to seeds having a defi- 

 nite wind-dispersal mechanism, or to seeds of extreme lightness. Of the 

 former the achenes of the compositse with their parachute pappus are the 

 best known instance,-— of the latter the seeds of Orchids. To take the 

 latter first it is by reason of their lightness that the seeds of Orchids are 

 able to reach and germinate in the branches of trees. But the distance 

 which such seeds are able to travel is probably slight, and this is borne out 

 Dy the extremely restricted distribution of the majority of the species. As 

 regards the achenes of compositte they can certainly travel great distances. 

 Recently experiments have been made to discover the air-movement re- 

 quired to keep various achenes upon the wing (so to speak) : (Proceedings 



