430 E. J. ALLEN. 



before with Thalassiosira. A good growth resulted, shoAmg t'lf-t the 

 animals do not excrete substances which completely inhibit the growth 

 of the diatoms. 



Ecaporated Tank Water. A number of experimients were made in 

 which a quantity of sea- water from the Laboratory tanks was evaporated 

 to dryness on a water bath, the residue heated to different degrees, 

 treated with strong, pure hydrochloric acid and evaporated two or three 

 times to get rid of the acid, and then redissolved to the original volume 

 in double-distilled water. After being neutralized by the addition of 

 NaHCO;., 4 per cent of the resulting solution was added to artificial 

 miqueled sea-water, the resulting culture medium being boiled, cooled 

 and inoculated with Thalassiosira in the usual way. 



The results of these experiments are set out in summary form in the 

 annexei Table A. In each case proper control experiments were set up 

 at the same time, generally one with artificial miqueled sea-water to 

 which nothing was added, and one with the same water to which 4 

 per cent of tank water was added, and the controls were boiled at the 

 same time as the other flasks of the experiment. 



As is seen from the table, five series of experiments were made. In the 

 first (Series A) the salts obtained by evaporating the ank \^ ater were 

 heated in a porcelain dish over a bunsen burner, the heating being carried 

 out carefully so that the flame did not actually touch the dish, which 

 never became anywhere near red hot. In Series B the evaporation and 

 heating were done in a platinum basin, which was raised to a dull red 

 heat over a bunsen. In Series C the salts were again evaporated and 

 heated in a porcelain basin and made as hot as they could be with a 

 bunsen burner, the flame of which played directly on the outside of the 

 dish, and was moved about so as to heat different portions in turn. In 

 Series D the salts were heated in a hot-air oven, being kept at a tempera- 

 ture of 164° to 170^' C. for an hour. In Series E the heating was again 

 carried out in a hot-air oven, a temperature of from 200' to 237° C. being 

 maintained for two hours. 



In Series A, D and E, where the heating of the residue was not excessive, 

 quite good cultures resulted. Although they did not quite come up to the 

 controls in which 4 per cent of tank water was added, they were in 

 every case altogether of a different order from what took place in the 

 controls in artificial miqueled sea-water to which nothing had been 

 added. 



In the other two series, B and C, where the degree of heating was much 

 greater, in most cases the culture was an entire failure, and in those 



