110 
ME. W. R. BOTJ8FIELI) : IONIC SIZE IN RELATION TO 
there appears to be no break in the values for the solutions. St is therefore calculated 
throughout from the value for water of 618-9. Sp is calculated from the densitv at 
1 per cent, by a linear formula. 
The series of observations set out in the two foregoing tables are plotted in Diagram 
I.? and shown by black circles. A series of observations given in a recent paper bv 
Gruneisen (‘Wiss. Abhand., Phys. Tech. Reichs.,’ 4, 239, 1905), to which fuller 
reference will later he made, is also plotted. They are indicated by double open 
circles, and the values are set out below. 
VI. 
Viscosity referred to water 
as unity. 
1 
1-0858 
0-5 
1-0410 
0-2 
1-0167 
o-i 
1-0086 
0-05 
1-0046 
It will be observed that in the more concentrated solutions there is a fair agree¬ 
ment. But at the dilute end of the curve the present observations show a consider¬ 
ably greater viscosity than Gruneisen’s observations. 
It hardly seems possible that the difference can he due to changes in the apparatus 
constant during the series, as the water constant is the same at the beginning and 
end. No othei senes so coherent as the one set out above was obtained, but the 
following observations, made with another instrument, tend to show that the difference 
is slightly greater than is indicated by the above series. 
Table IV.—Other Determinations of Viscosity of Dilute NaCl solutions. 
VI. 
Viscosity referred to water 
as unity. 
0-00085 
1-0025 
0-0090 
1-0050 
0-0120 
1-0054 
0-01563 
1-0054 
0-0286 
1-0070 
0-0343 
1-0064 
0-0491 
1-0077 
these determinations have only the value of isolated observations, but they are 
entitled to some weight, and they are indicated on the diagram by small open circles. 
1 he cui\ e drawn on the diagram indicates the line of values actually adopted for the 
