138 SPOLIA ZEYLANICA. 



into the lake. The oysters average 1-778 x 1 " 946 inch, 

 against 2*04 X 2*16 inches in May, 1912. They are very 

 abundant on the whole. 



Area S is completely covered with oysters of very good 

 average size, and very abundant. They average 2*094 x 

 2 '354 inches. In May, 1912, only one specimen measuring 

 2 • 87 x 3 • 25 inches was taken in this area. 



Area T is another well-stocked and satisfactory area, the 

 oysters being both abundant and of good average size. They 

 average 2-012 X 2-274 inches. 



Area U. — Oysters abundant in the south-east portion of the 

 area, but scanty further north. The}^ average 2-08 x 2-391 

 inches. 



Area V. — Oysters abundant and having an average size of 

 1-87 X 2 -162 inches. 



Area W. — These three last areas were quite barren of oysters 

 last year and the year before, and this lack of oysters was pu4i» 

 down to the extreme softness and thickness of the mud which 

 forms the bottom of these areas, but apparently this is no 

 preventative of oysters in their j^oung stages at least. The 

 average size for this area was 1 - 766 X 2-075 inches. 



It will be seen that the deposit of oysters this year, while 

 more extensive and evenly distributed than last year or the 

 year before, consists of smaller oj^sters on the whole. Also 

 one does not find the densely packed patches which were 

 evident last year. One would naturally expect as a result 

 of this better distribution, a corresponding increase of size 

 compared with last year's oysters, but this is not so. It is 

 possible that the present bed of oysters was spawned later in 

 the year than the 1912 bed, and this seems the most probable 

 explanation, because the rains (which almost certainly prevent 

 spawning while they last) continued much later this year than 

 they did in 1912. 



In addition to the work of inspecting, a series of water- 

 samples was taken in various parts of the lake for the purpose 

 of determining the salinity and its bearing on the questions of 

 distribution and growth-rate, but the samples have not yet 

 been tested. 



