24 THE AMERICAN MONTHLY [January, 



admirably fixed specimens shrivelling will often appear when al- 

 cohol is applied. This difficulty is partly overcome, with great 

 pains, by using a series of alcohols of ascending degrees of 

 strength. But the result of this mode of procedure is not by any 

 means always satisfactory. 



Dr. Cobb has described a method by which, in the case of 

 small organisms, the shrinkage due to change from one fluid to 

 another of ditierent density may be reduced to a minimum. In 

 his ditierentiator we have an instrument of admirable simplicity 

 for insuring this result. But I have found that in practice the 

 use of the difterentiator involves a considerable expenditure of 

 time. To get a specimen from distilled water to 90 per cent, 

 solution of alcohol, for example, no fewer than eleven diflerent 

 mixtures of water and alcohol have to be made up and poured 

 into the reservoir tube. 



A simple piece of apparatus which I have devised does away 

 entirely with this — the gradual substitution one for another of the 

 two fluids of diflerent densities being effected automatically. An 

 obvious mode of meeting the difficulty suggests itself at once. 

 Why not have the second fluid falling into the first drop by drop, 

 mixing thus very gradually with it and eventually replacing it.'' 

 The difficulty in the way of this is that as each drop of the much 

 lighter liquid enters the denser, violent though circumscribed 

 currents are produced which are damaging to the delicate organ- 

 isms we ai'e dealing with. 



The requisites for the method about to be described are — sev- 

 eral reservoirs of glass or earthenware fitted with glass taps and 

 having each a capacity of a gallon or more, some wide-mouthed 

 bottles of a variety of sizes, fitted with perforated india-rubber 

 stoppers, and some lengths of glass and india-rubber tubing. 



Two bottles of similar size are connected together by tubing 

 in the way represented in the figure. One of these. A, we call 

 the mixing bottle; the other, B, contains the objects, and must 

 have a capacity equal to at least a hundred times the bulk of the 

 latter. The objects are in fluid i, and it is desired to substitute 

 fluid 3. Both bottles are filled, or partially filled, according to 

 circumstances, with fluid i, and bottle A is connected with the 

 reservoir of fluid 2. It is somewhat difficult by means of u tap 

 to regulate the flow so that, let us say, one drop in five seconds 

 will pass out of the reservoir ; and it is much more convenient 

 to effect this by intercalating in the supply pipe a section of 

 glass tubing drawn out to the required degree of fineness (repre- 

 sented in the figure as disconnected from the proximal portion of 

 the supply tube). The rate of flow through this narrow section 

 of the tube can be further regulated by raising or lowering the 

 reservoir or the mixing bottle, thus alternating the pressure. 

 With bottle B is connected an overflow tube. Above the narrow 

 section of glass tubing in the supply pipe it is well to have a 

 piece of filter paper stretched across the mouth of the piece of 



