BETWEEN THE FREEZING AND BOILING-POINTS. 
151 
so much was due to his kind supervision in perfecting different portions of the 
apparatus. Mr. Stovel also was obliged to leave at this time, so that the work was 
somewhat delayed. 
During the summer of 1898, Mr. Charles Sheffield, B.Sc., was kind enough to 
devote his entire time to the work, and made himself exceedingly useful through his 
untiring efforts until late in the autumn. University duties being closed, it was 
possible for us to devote all our time to the work. The measurement of the 
mechanical equivalent we obtained will be described 'further on, but it must be 
regarded as a preliminary attempt Owing to a source of error in the method, which 
was not discovered until some time after. We made a careful study of the general 
theory of the method, which, as will be explained in its place, was affected somewhat 
by the error above mentioned. On comparing these determinations with later ones, 
and more particularly in applying the theory of the method to different calorimeters, 
we met with such large discrepancies, much larger than any possible error in the 
instrumental readings, that we were forced to abandon the greater part of our earlier 
results, and re-organize the experiment. 
Unfortunately Mr. Sheffield was called away at this time and was unable to 
continue his work on the method. It is with extreme regret that I have to record 
the death of Mr. Sheffield, since leaving this laboratory, which occurred recently at 
Niagara Falls, where he occupied a position in the capacity of electrical engineer. 
His death at so early an age and under such trying circumstances is all the more sad, 
as he had proved his worth and ability in so many ways as an accurate observer and 
faithful worker. 
During the winter of 1898-99 I was obliged to undertake the sole responsibility 
s 
of the work, with the exception of some temporary assistance in taking observations 
from my colleague, Mr. H. M. Tory, M.A., to whom I am also indebted for many 
helpful suggestions. As at that time there was no prospect, until the close of the 
session, of finding an experienced observer who could devote sufficient time to help 
in taking observations, it was necessary for me to arrange the. conditions of the 
experiment so as to be able to take all the observations, both thermal and electrical, 
myself. With a little practice I was soon satisfied that this could be done, although 
not quite so quickly as with two observers, yet with sufficient accuracy to satisfy the 
conditions of the experiment. It became chiefly necessary to produce perfectly 
steady and uniform conditions, over a more extended period of time, conditions which 
demanded greater refinements in the apparatus. The experiments from this time on, 
as they became little by little improved and extended, were so steady and consistent, 
and fulfilled the conditions demanded by the theory so perfectly, that it was deemed 
unnecessary to break the continuity of the work by introducing a second observer. 
From January, 1899, to the close of the work, the complete set of observations for 
nearly every experiment was taken by one observer. 
The results of the work from 4° C. to 60° C., obtained between January and June 
