CHEMISTR Y: HARKINS AND HUMPHER Y 
587 
represent the values of the function of r/a, and the abscissae give the 
values of r/a. The complete circles represent Lohnstein's theoretically 
determined values, while the circles given in outline represent the new 
values determined experimentally. The figure shows that the experi- 
mental values give with Lohnstein's first two points for small values of 
r/a a much smoother curve than is given by his own theoretical values. 
The experimental values were determined by the use of a number of 
different liquids, and measurements were made both upon liquid inter- 
faces and at the surface of a single liquid. 
Morgan, who has carried out an extensive series of investigations^ 
on the drop-weight method, does not use the general Lohnstein rela- 
TABLE I. 
Experimental determination of the values of the function f (r/a) 
Point 
Temperature 
No. 
r/a 
U>/a) 
Interface 
degrea 
1 
0.281 
0.709 
Water: Ethyl Carbonate 
25 
2 
0.366 
0.685 
Water: Benzene 
25 
3 
0.441 
0.672 
Water: Dimethylaniline 
25 
4 
0.484 
0.654 
Water: Ethyl Carbonate 
25 
5 
0.592 
0.639 
Water: Benzene 
10 
6 
0.621 
0.636 
Water: Benzene 
20 
7 
0.633 
0.632 
Water: Xylene 
25 
8 
0.636 
0.634 
Water: Benzene 
25 
9 
0.648 
0.634 
Water: Benzene 
30 
10 
0.649 
0.632 
Water: Toluene 
25 
11 
0.709 
0.620 
Water: Air 
25 
12 
0.837 
0.615 
Water: Hexane 
25 
13 
0.845 
0.616 
Aqueous solution of Sodium 
25 
Chloride: Benzene 
25 
14 
1.071 
0.612 
Benzene: Air 
25 
15 
1.387 
0.620 
Aqueous Solution of Strontium 
Bromide: Hexane 
25 
tion, but considers that the law of Tate holds for drops which have a 
'normal' form. This law he expresses in the form: 'Surface-tension = 
Constant X Drop Weight.' The use of this equation is equivalent to 
the assumption that in figure 1 the curve at the bottom is coincident 
with its horizontal tangent. From the form of the curve it may be seen 
that this is not strictly true at any point, but that no serious error is 
involved in this rule if the determinations of surface-tension are always 
made with a tip which gives a value of r/a very nearly that at which the 
tangent touches the curve. Usually the tube used does not meet this 
very specific requirement, and the result is therefore different from the 
true result by the distance between this tangent and the curve. 
The results of Morgan and McCann^ upon 5 different liquids with 
16 different tips have been used to calculate the values of the function 
