THE CULTIVATION OF DIATOMS. 149 



these delicate cultivations, the experimenter will never regret that 

 he has used the greatest parsimony in adding organic substances. 



I have before said how these sowings should be made, so 

 that I need not return to that part of the subject ; but I ought to 

 tell how you may, and ought, to procure Marine Diatoms, for 

 charging these solutions. 



By immersing fresh oyster shells that have been very carefully 

 deprived of all traces of the flesh of the mollusc, so as to avoid 

 all putrefaction, in sea-water, either natural or artificial, you form 

 a growth that can be carried on for one or two months, and pro- 

 duce often fine specimens of Diatoms, that you can further 

 cultivate at leisure. The oysters should be placed in stone-ware 

 or earthen vessels, which by their opacity scatter the reflections 

 of the sun and wafls, and only allow the light that proceeds from 

 near the zenith to reach the Diatoms. 



If a correspondent is caUed on to send away Diatoms, he 

 should, in the first place, separate by decantation all muddy 

 substances, and after washing them with clean sea-water five or 

 six times, enclose them in a tight vessel with a large excess of 

 sea-water. By adopting this plan, living marine Diatoms may be 

 kept several weeks, whilst if Diatoms be sent with the sediment 

 from which they have been gathered, putrefaction will kill them 

 in a few days. 



When the growth of marine species has to be continued for a 

 long time, it is indispensable to prevent the sea-water from alter- 

 ing its density, etc., by concentration. Many forms of apparatus 

 have been proposed for preventing the evaporation of the water, 

 and the lowering of the level w^hich results from it ; that which I 

 represent in Fig. i, PI. VIII., appears to me the most practicable 

 and trustworthy. 



Fig. I. — Fis a vessel containing growing Diatoms, filled with 

 sea-water. 7^ is a flask containing distilled water, to maintain 

 a constant level. S' is a siphon which conducts the water, and 

 6" a tube that permits the action of the siphon when the level of 

 the liquid in F falls below its orifice, and when, consequently, the 

 air in the flask communicates freely with the external air. The 

 theory of this intermittent siphon is too simple to require any 

 explanation to my readers. 



