between the viscosity of liquids and their chemical nature. 689 
Consequently, 
rj = p 
where r] is expressed in terms of the slope. 
Let r]x^, rjB,, <^ 0 ., be the viscosity coelBcients of the liquids A, B, C, &c., measured 
at the slope S^, and let rjc^, &c., be the corresponding coefficients at the slope 
Then, if the relations betwmen 77 v,; &c., are the same as those betwmen 
’?B,? Vo.:> i is, if the relations are to be independent of the value of the 
slope, it is evident that the ratios 
VaJva,, VbJvb,, VcJvc,, &c., must be equal. 
But 77 a^, &c., can be expressed in terms of the slope S^, and may be written 
and 17 A,, &c., can be in 
written 
+ + &c., 
tlie same way expressed in 
terms of the slope S 3 and 
and hence, if the relationships between &c., are to be the same as those 
between 77 a^, 773^, 770^, &c., it follows that the values of ratios 
( 81 / 83 )"^^'''^ ( 82 / 83 )’'®''^”®'^ &c,, must be equal, 
and since in comparisons at equal slopes and S 3 have the same value for all the 
liquids, it follows that the comparisons will be independent of the slope if 
^^a/('«a + 1 ) = + 1 ) = &c. 
That is, if Wa = = nc, &c. 
So far as Slotte’s formula goes, it is thus indicated that for the comparisons to be 
general the value of n should be the same for all the liquids. 
It has already been stated that, from the mode in which n is deduced, its value is 
affected by circumstances mmre or less accidental. It is satisfactoiy, however, to find, 
on comparing the values of n for all the liquids, with the exception of aldehyde and 
the alcohols, that the variation is comparatively small. 
The following table contains the mean value of n as given by the difierent series of 
liquids, the alcohols and aldehyde excluded ;— 
4 T 
MDCCCXCIV.—A. 
