C. F. ROUSSELET ON SOME MICRO-CEMENTS FOR FLUID CELLS. 97 



penetrating glycerine would soon have found its way through. 

 This is the general principle, varying the varnishes as circum- 

 stances and the mounting fluid require. Marine glue, much in 

 vogue for securing cells to glass, has failed to stand the test of 

 time ; india-rubber compounds are notoriously unstable. For 

 fixing glass cells on slips I have made a firm joint with a mixture 

 of red lead and white lead after the manner of the blacksmith's 

 hot-water joint. I take flake white from the tube and work 

 powdered red lead into it with a palette knife until it becomes too 

 stiff to work comfortably ; I then dilute with gold size only a 

 small portion at a time as I want it, as it hardens very rapidly 

 when the gold size is added. Press the cell hard down so as to 

 have the thinnest secure film of cement, and keep in a warm place 

 for a few days ; turn off the surplus ooze before it has become too 

 hard, and do not use the cell for at least a fortnight. Your 

 excavated cells and solid cast ones used by Hett are of course 

 better than any built-up ones." 



It will be noticed that Mr. Suffolk advocates five or six thin 

 coats of gold size instead of one thick one. The reason is that oil 

 varnishes, after the evaporation of the turpentine, harden not by 

 evaporation of the oil, but by its oxidation, forming a tough layer, 

 and therefore the process goes on best in a thin layer. 



After carefully considering all that I have learned by Mr. 

 Sufiblk's letters, I would recommend for extreme durability for 

 fluid mounts such as mine, in which the mounting fluid is forma- 

 lin, or similar watery fluids, first a coat of pure damar in benzole, 

 then a coat of the mixture of damar and gold size, followed by 

 three or four thin coats of pure gold size at intervals of twenty- 

 four hours, and lastly a finishing coat of Ward's brown cement. 

 The "extra stout" gold size I am inclined to discard again, 

 because it does not adhere to glass so firmly as the ordinary 

 quahty. 



JouRN. Q. M. C, Series II., No. 43. 



