392 Proceedings of Royal Society of Edinburgh. [sess. 
course water should be brought to the thermometer in such a way 
as not to be subject to a change of temperature in the meanwhile. 
The determination of the specific gravity of water was for a 
long time considered as a delicate piece of work, which could be 
done only by the hands of the scientist. It was very brave of me 
to dare to take in hand the hydrometer, but I became in a short 
time master of this ordinary sort of work. After taking charge of 
this operation in 1886, I, being a captain then, handed it to all my 
officers on board, and was really glad to see that every one of them 
was as particular about the exactness of the observations as a man 
who is accustomed to do this kind of work. It may be interesting 
to mention that out of ten instruments given to the officers of the 
watch only two were broken after six thousand observations had 
been made during the three years’ voyage uninterruptedly in good 
weather and bad. The association of all the officers in this work 
gave me very great assistance. When the work is done by a 
scientist, observations of specific gravity may be taken once or 
several times a day, but, as every officer of the watch could do it 
on board my ship, I sometimes had observations made every five 
minutes. It is not necessary in ordinary circumstances to de- 
termine the specific gravity of the water every five minutes, but 
if you cross a certain current and wish to know the limits of it, 
you have to increase the number of observations. 
I wish to draw your attention to one fact, which should not 
be neglected. The hydrometer changes its position, and conse- 
quently its reading, when you place the thermometer in the same 
vessel. I found out this fact and published it in my book 
Le Vitiaz et VOcean pacifique (paragraph 58). Dr Krummel, 
of Kiel, - afterwards studied this subject particularly. I mention 
this fact because it is better not to have the thermometer in a 
vessel in which the hydrometer is floating. In any case, observa- 
tions of the specific gravity should be done in the same way in 
which the errors of the hydrometer were determined. 
I will now show you the results of my work done at different 
times, from 1881 to 1889. My books and papers are published 
chiefly in the Kussian language, and for this reason they are not 
very well known in this country. I cannot, in the course of my 
address, make you familiar with all my works, and wish only to 
