154 



PALMER : 



corrosive sublimate. Into a boiling saturated watery solution 

 of this salt, contained in a test-tube, a quantity of the fresh 

 diatoms was dropped as quickly as possible from a pipette. 

 The fixing was continued only a few minutes, and the diatoms 

 were then washed free of the chemical upon a filter kept con- 

 stantly^ full of water. The thoroughly washed filter and its 

 contents were now covered with water in a large watch-glass, 

 and the diatoms properly stained. Several processes of stain- 

 ing were tried, but the best results, so far as concerns Siiri- 

 rella, were obtained firstly with aqueous eosin, and secondly 



■-TK 



FiCtURK III. 



with iron-alum and haematoxylin. Those stained with eosin 

 were mounted in 2 per cent, formalin cells, and the iron-alum- 

 hsematoxylin stainings were, with the greatest care, deh)'- 

 drated with alcohol and put through oil of cloves into balsam. 

 This dehj'dration and final transfer required nearly a week of 

 waiting. The balsam mounts showed very heavy stainings 

 in the body of the cell, and precisely the right amount of 

 color in the ribs and keels of Surinlla. Moreover, the pro- 

 cess of fixing had been exceedingly satisfactory as regards 

 avoidance of excessive shrinking of the protoplasm. In 

 Figure III I give a sketch of the aspect of affairs in the keel 



