REPORT ON THE COMPOSITION OE OCEAN-WATER. 
143 
Method of Gas Analysis. 
The volume of sea-water gas in each tube, when measured at the ordinary tempera- 
ture and pressure, generally amounted to some 15 c.c. ; sometimes it was more, but more 
frequently it was less. 
With such small samples, and especially considering the impossibility of replacing 
them, it would have been imprudent to attempt anything beyond the determination of 
what were presumably their substantial components. It was also clear to me from the 
first that in such a case substantial correctness and reliability in the numerical results is 
worth more than high but unguaranteed precision. Therefore, small as the samples 
were, I decided upon dividing each into two approximately equal parts, and analysing 
each separately. Of the several kinds of gas-analysis apparatus which have been 
invented, Doyere’s seemed to me to be the one which would probably best adapt itself 
to my requirements. But I had not this apparatus in my possession, and to procure one 
from Paris would have led to considerable loss of time. Besides, a few experiments which, 
by the kindness of my friend Dr. Bonalds of Bonnington, I had been enabled to make 
with one in Iris private laboratory, had revealed to me certain difficulties in its manage- 
ment, which made me shrink from its unqualified adoption, although I felt that they 
might be purely subjective and conquerable by long practice. At the time I fortunately 
commanded, in the person of Mr. Robert Lennox, the services of an assistant who, 
besides being an excellent experimenter generally, is an accomplished glass-blower ; and 
with his help I thought I might try to construct a modified and improved kind of 
Doyere’s apparatus on my own premises. Our joint efforts ultimately resulted in the 
apparatus represented on PI. II., which, I may say at once, was found to work very 
satisfactorily, and served for all the gasometric work to be here reported on. 
The apparatus, apart from the two mercurial troughs required, consists of three 
separate parts, namely — a “ measurer,” an “ exploder,” and a “ pipette ” for the absorp- 
tions. 
The measurer, as shown by the figure, is a combination of a wide with a narrow glass 
tube, after the manner of a Gay-Lussac burette. The narrow tube is soldered into the 
side of the wide one somewhere near its bottom, and is bent up so as to run on along- 
side of and lie flat against it. The wide tube by its lower contracted end communicates 
with a long capillary tube of vulcanised india-rubber, and through it with a mercury- 
reservoir like a Geissler air-pump. At their upper ends the narrow and the wide tube 
are both provided with good Geissler stop-cocks ; to the exit end of the one at the wide 
tube is soldered a capillary JJ-tube, similar to the one characteristic of Ettling’s gas- 
pipette. The wide tube bears a millimetre-scale to read off the position of the mercury 
