108 
MB. W. B. BOUSFIELD: IONIC SIZE IN BELATION TO 
The differential formula becomes on substituting for 77 , p, A, B and t, 
8 V = 0-01053 op + 0-000018406 St. 
It was found by trial with a series of water values at different temperatures that 
the differential formula remained sufficiently accurate even when St amounted to as 
much as 50 seconds. 
I he values determined lor the less dilute solutions of NaCl are given in the 
following table :— 
Table I.—Density and Viscosity of NaCl Solutions at 18° C. 
Viscosity referred 
1. 
m. 
Density. 
Viscosity. 
to water as 
unity. 
6 
1-068 
1-04200 
0-01150 
1-093 
4 
0-7024 
1-02749 
0-01112 
1 • 058 
3 
0-5232 
1-02027 
0-01098 
1-042 
2 
0-3463 
1-01308 
0-01083 
1-030 
1 
0-1720 
1-00592 
0-01068 
1-016 
0 
0 
0-99866 
0-010514 
1-000 
Ihe duplicate density determinations in no case differed by more than a unit in the 
fifth place of decimals. The air corrections were determined from observation of 
baiometer pressure and laboratory temperature, and it is considered that the densities 
may be taken to be relatively correct within ± 1 in the fifth place of decimals. The 
viscosity determinations have not the same order of accuracy. 
hoi the purposes of this paper it was a matter of considerable importance to 
determine the viscosities of more dilute solutions, but this proved to be a matter of 
the greatest difficulty. Other observers of viscosities have had the same experience, 
but whether owing to the same cause is not clear. 
In the former paper it was- pointed out that the constants of the apparatus, as 
tested by the time of flow of water at 18" C., vary from time to time in an irregular 
manner. No amount of washing will bring the apparatus to a steady condition. I 11 
dealing with very dilute solutions the variations of the water constant are often of 
the same order of magnitude as the variations due to variation of concentration of the 
solution. Hence the only safe method of operation is to interpose between the 
observations of salt solutions observations of water, which, of course, involves a 
thorough rinsing of the apparatus, which may in itself introduce some change in 
its condition. 
When a series of observations of solutions with interposed observations of water 
shows no material change 111 the water constant, it is assumed that the conditions 
were the same throughout, and that the observations are therefore accordant. But 
this assumption is not always true. A very large number of observations on dilute 
