CEYLON PKART. BANKS. 217 



recognizable on the Tuticorin banks and on the Ceylon banks, and 

 which finds an exit through the Paumben Channel. Such a surface 

 drift would be sufficient to produce the results which the drift 

 experiments have shown. 



The second explanation is, that when the cuiiiMit strikes Cape 

 Couiorin a small portion of it is deflected uj) the Indian side of the 

 Gulf. If the main stream strikes the Ceylon coast near Colombo 

 it will flow southwards, l)ut a small portion may he deHected 

 northwards into the Gulf of Mannar. 



Messrs. Southwell and Kerkham have very properly confined 

 themselves to the results of their drift-bottle experiments, but in so 

 doing they have not fully discussed the causes Avhich produced those 

 results. For instance, they speak constantly of "currents," and 

 have evidently not realized that the drift experiments do not assist 

 them in discriminatmg between oceanic currents and the ordinary' 

 surface drift caused by the wind , except in cases where the current 

 and wind are in opposition. In a weak south-west monsoon, for 

 instance, they speak of the northerly currents floAving along the 

 Ceylon and Indian sides of the Gulf of Mannar. It is just as likety 

 that at that phase of the monsoon the Gulf is only affected by surface 

 drift. Drift-bottle experiments, however, cannot discriminate 

 between the one and the other, and before we have any right to 

 speak of oceanic currents affecting the surface waters of the Gulf 

 of Mannar, we must supplement the very valuable results obtained 

 from Southwell's and Kerkham's drift-bottle experiments by the 

 chemical examination of a large series of water samples from the 

 area under discussion. Only bj^ this means will it be possible to 

 solve the current questions in the Gulf of Mannar. 



Bearing of Drift-bottle Experiments. 



The bearing which these drift-bottle experiments have upon the 

 main question is probably a very intimate one. In the Gulf of 

 Mannar there are two series of pearl banks : those on the Indian side 

 at Tuticorin, and those on the Ceylon side ; the two are separated 

 by a distance of nearly one hundred miles and by a considerable 

 depth of water. It is highly probable that after a series of barren 

 years, as we are passing through at present, the Ceylon banks are 

 replenished through the agency of the Tuticorin beds. 



As we have akeady seen, during a very strong south- v/est monsoon 

 the oceanic current sweeps upas far as the Tuticorin beds, and then 

 takes an eastward course as far as the pearl banks. One of the 

 8pa^\Tling maxima coincides with the early part of the south-west 

 monsoon, so that it is possible that the floating larvae liberated on 

 the Tuticorin banks may be carried over to the Ceylon side if thej' 

 happen to be in the way of the oceanic current. According to 

 Homell the pelagic stage lasts at least five days. Southwell and 

 Kerkham have shown how in the case of a strong monsoon the 



