BY N. A. COBB. 161 



will be found in the rack now to be described. A strip of wood, 

 whose length must be regulated by the number of differentiators it 

 is desired to use, has driven through it, at intervals of from 

 twelve to eighteen inches, long wire nails, whose points will then 

 project an inch. The strip of wood so prepared is fastened in a 

 horizontal position with the nail-points also in a horizontal 

 position. Take a large cork, perforated to fit, and slide it on to 

 the differentiator reservoir. Move the cork back and forth until 

 the whole instrument will balance when left free with the cork 

 resting on the fore-finger — in short, place the cork opposite the 

 centre of gravity. If the cork in this position be impaled upon 

 one of the nails, the latter will be found to afford a good pivot 

 for all movements necessary in regulating the flow. 



I formerly gave the long arm of the filter the same diameter as 

 the short arm. Later I found it better to make the filter in the 

 form shown in Fig. 3, i.e., with the long arm semicapillary. 

 When the diameter of this part of the filter was large, space was 

 given for the mixing of the fluids which passed into it, the order of 

 the liquids obviously having a tendency to become reversed after 

 passing the u-bend of the filter. This gave rise to two difficulties. 

 First, when the filter was reversed on changing to the alternate 

 reservoir, a small backward diffusion took place which was especially 

 inconvenient after dehydrating with absolute alcohol ; secondly, 

 the precipitate which occurs when carmine fluids are mixed with 

 alcohols had a tendency to accumulate in a flocculent condition 

 in the long arm of the filter, a fact due no doubt to the diffusion 

 constantly taking place there, and this precipitate sometimes 

 clogged the orifice. Both difficulties are avoided in the new form. 



The remedy for clogged orifice may as well be given here. If 



only the extremity is clogged, break away a little of the capillary 



part, which should always be long to allow for this emergency, 



and readjust the flow if necessary. If that procedure be inad- 



missable, hold the point of the filter for a fraction of a second in 



a small alcohol flame. The slight explosion which occurs will 



often I'emove the obstruction. 

 11 



