The Microscope. 351 



methods, he found that precipitation of mercurous sulphide in 

 situ was the best method of filling these cavities. If clean 

 Diatoms are soaked in a solution of subnitrate of mercury till 

 all their hollows are filled with it, and if they are then im- 

 mersed in sulphide of ammonium, a double decomposition takes 

 place, whereby black insoluble sulphide of mercury is produced 

 and left in the minute cavities in wliich it is formed. Careful 

 levigatiion with water will free the charged Diatoms from the 

 greater part of the loose and unconfined sulphide, and leave 

 them clean enough for examination. The process shows that in 

 Diatoms of all classes the markings, whether secondary or pri- 

 mar}^ are due to the existence of holes or cavities in the sub- 

 stance of the test. Even the finest of the secondary markings 

 can be filled with an opaque material. 



Mr. Gill also writes to the Journ. R. Micros. Soc. on the same 

 subject : When cleaned and dry Diatons are soaked in a con- 

 centrated solution of perchloride of iron for some time, all hol- 

 low spaces contained in the frustnles become charged with the 

 ii'on salt. If they be now transferred to an acid solution of 

 potassium ferrocyanide, Prussian blue will be formed both out- 

 side and inside all hollows and cavities. On washing and levi- 

 gating with water, the outside, unconfined portion of the precipi- 

 tate can be washed away in great part, while those portions 

 which are more or less surrounded by walls of silica remain in 

 place, and serve to mark clearly the positions and limits of the 

 spaces containing them. 



Evaporating a solution of sodium platinum chloride on 

 cleansed Diatoms, and igniting the whole with the addition of 

 some crystals of oxalic acid, serves to charge the minute cavi- 

 ties with a deposit of spongy platinum. 



Pinnulariee under either of these treatments show their 

 coarse ribbing to consist of ribbon-shaped tubes contained in 

 the walls of the frustule. Fleurosigma, Stauroneis., Gocconema, 

 etc., show their dots to be spaces which can be filled with 

 foreign bodies. Goscinodisci have the openings into their 

 lacunae so large that the precipitates, for the most part, are 

 washed out in the course of mounting, but the cell walls take so 

 much of color that their shape and parts can be clearly dis- 

 tinguished. 



