18 THE VOYAGE OF H.M.S. CHALLENGER. 



commence the forcing of mercury into the small bulb. The mercury forced in was 

 weighed ; then the capacity of the small bulb was determined by weighing its content in 

 mercury. The difference of these weights is the weight of mercury, which would occupy 

 the same volume as did the air when compressed. Finally, the original volume of the 

 air was found by weighing the whole apparatus, first empty then filled with water ; and, 

 most important in view of Amagat's results, the barometer and thermometer were care- 

 fully observed at the instant when the apparatus had its lower end placed in the vessel 

 of mercury. Mr Kemp, who made these instruments for me, suggested and carried out 

 the great improvement of inserting a small triangular pyramid of glass into the choked 

 part of the bore (as shown in the small sketch). The effect is to break the mercury 

 (which must be very clean) into exceedingly small drops. In this way the actual 

 compression of the air was determined with a limit of error, represented at the utmost by 

 the ratio of the volume of one of the small drops of mercury formed at the obstruction to 

 the whole capacity of the small bulb. By working simultaneously with three instru- 

 ments of this kind, even this very small error could be in great part eliminated : — and, 

 practically, the compressions were measured far more accurately than was at all necessary 

 for the purpose in hand. For greater accuracy a larger apparatus would be required. 

 This, however, was quite unnecessary. And the requisite limit of accuracy in the experi- 

 ment rendered it unnecessary to correct for the alteration of volume of the smaller bulb 

 consequent on the pressure to which it was subjected. 



In my later experiments a long carefully-gauged tube of l - 5 mm. in bore was sub- 

 stituted for the small bull). This tube was coated internally with an excessively thin 

 film of metallic silver thrown down by sugar of milk. The process was arrested the 

 moment the film became visible by reflection. This film is at once dissolved by the 

 mercury up to the point which it reaches at the greatest pressure, and leaves a perfectly 

 sharp and nearly opaque edge from which to measure. This device has proved so very suc- 

 cessful that I have now substituted it for the indices in all the pressure gauges (shortly to 

 be described) which are employed for very accurate measures. And I am at present engaged 

 in measuring, by comparison of a glass gauge and an air gauge both fitted in this manner, 

 the compression of various gases at pressures up to fourfold those applied by Amagat, 



Internal Pressure Gauges. 



The next step was to find some plan of construction for an instrument which, having 

 its scale determined once for all by comparison with the air-gauge, should ever afterwards 

 serve instead of it, thus affording a ready measure of pressure. Liquids are obviously 

 better fitted for this purpose than solids, if only on account of their absolute homogeneity 

 and their greater compressibility. But, unfortunately, two liquids must be employed, 

 since a record must be kept : — the apparatus being surrounded on all sides by 9 inches 

 of iron : — and, as will be seen in Appendix E, all my trials with two liquids were more 



