436 E. J. ALLEN. 



also made no appreciable difference. Even when the dilution was ex- 

 tended to 200 per cent a fair quantity of growth took place. 



The following experiment was made in order to test the same point on 

 Thalassiosira gravida. 



Experimetit 476. Artificial sea-water was made up with the normal 

 relative proportions of salts, but of double the normal strength. A series 

 of dilutions was then prepared, doubly distilled water being added in the 

 proportions stated : 



The right quantities of Miquel's solutions were added to each, and 

 4 per cent of sea-water from the Laboratory tanks. Flasks were then 

 inoculated with three drops each of Thalassiosira gravida culture. No 

 growth took place in A and B. Excellent, healthy growths with good 

 chain formation took place in all the others. E and F were best, and one as 

 good as the other. G and D were excellent growths, but the quantity at 

 any time was less than in E and F. In C, H and J, although the growths 

 were quite good the quantity was considerably less than in E and F, that 

 in C also being less than in H and J. 



It will thus be seen that very considerable changes in the salinity of the 

 culture medium can be made without much effect being produced on the 

 growth of Thalassiosira. Dilution of the medium is less detrimental than 

 concentration. 



The experiments described in this section show how wide the 

 variation in the chemical composition of the culture medium may be 

 without any very marked effect being produced on the growth of the 

 diatoms. The difficulty in growing the diatoms in artificial sea-water is 

 clearly not due, as at one time I thought might be the case, to the fact that 

 a very delicate balance between the amounts of the different salts is 



