Tensions of Liquids and their Chemical Constitution. 87 



up through two right angles, the orifice being exactly in the 

 axis of the cylinder, no matter how the pipette was inserted. 

 It is of some importance that the orifice be always placed in 

 such a position, that the drops may be discharged vertically. 

 A second pipette with a straight stem, so as to discharge bub- 

 bles downwards, was also employed. To make a determination, 

 the 'pipette (the curved-stemmed one, if the dropping liquid is 

 the lighter, and the straight-stemmed one, if the dropping 

 liquid is the heavier) is filled with one of the liquids by suc- 

 tion through a long rubber tube slipped over the upper end of 

 the pipette and provided with a pinch cock. It is then 

 inserted in the cylinder which has previously been filled with 

 enough of the other liquid to stand an inch or so above the 

 orifice of the pipette. The free end of the rubber tubing is 

 slipped over a piece of glass tubing, so shaped that it can be 

 held conveniently in the hand and closed with the thumb. 

 The thumb is now kept tightly pressed over the tube, and the 

 pinchcock opened. A few drops fall rapidly at first, before 

 the pressure becomes sufficiently reduced. By lifting the 

 thumb, air may be admitted and the liquid made to drop at 

 any desired rate. After a little practice, the raising of the 

 thumb and the counting of the drops are done almost instinc- 

 tively. If the time of the growth of the drops is to be noted, 

 a watch is placed at the side of the apparatus, and the falling 

 of the drops made to coincide with the movements of the 

 second-hand. This is very easily done. Suppose, for instance, 

 one wants the drops to form at the rate of one a second. 

 Then, by beginning to count when the second-hand crosses a 

 quarter-minute mark, fifteen drops should have fallen when 

 the second-hand passes the next quarter-minute mark, thirty 

 drops, when it passes the next, and so on. In this way, with 

 very little effort, a strict account of the time taken by the 

 drops in forming can be kept. The number of drops, into 

 which the volume comprised between the two marks divided 

 was taken in every case. If the last drop before the mark was 

 reached did not leave the residual liquid nearly flush with the 

 mark, which was generally the case, another drop was allowed 

 to form, and, by observing the volume it occupied in the stem 

 of the pipette, an estimation to a quarter of a drop could be 

 made, which was added to the number of drops counted. 

 When one determination had been made, the pipette was 

 removed, wiped free from the outside liquid, refilled and again 

 inserted in the cylinder. The liquids in the outside vessel are 

 easily drawn off separately by means of the stop-cock. 



In regard to the accuracy of results obtained by this appa- 

 ratus, it was found that determinations of drops forming at 

 the same rate and dropping upwards, seldom differed, if the 



