Donovan — On the Comparable Self-acting Hygrometer. 559 
It is easy to know liow many rounds of the dial the primary index 
has performed, during the absence of an observer, by the indication of 
the secondary : the first digit of the decade last passed or now passing 
over, by the secondary index is the number of rounds already tra- 
versed by the primary, the constituent moisture of which is still 
present in the gut-line ; and this digit, along with the number shown 
by the primary index at the same time, gives at a glance the moisture 
of the atmosphere at that present moment ; and also the fraction of 
saturation, the denominator being 1000. Thus, on returning, after 
an absence, if I find the secondary index perhaps at or about 60° and 
the primary at 50°, I learn that the primary index has completed six 
rounds and a half; and that the constituent quantity of water, still 
present in the line, should be taken into account in the next obser- 
vations, or else removed by exsiccation. 
In noting the performance of the secondary index, its distance on 
the gut-line from the point where the latter is confined by the clamp- 
screw, must be also noted, that being an essential element of the 
observation. When the secondary index was attached to the gut- 
line, exactly midway between the apex of the cone and the clamp- 
screw, the primary index completed two rounds, while the secondary 
completed one, and both arrived at zero at the same time, they having 
started from that point. 
It is often necessary, when making hygrometric experiments, to 
subject the hygrometer to an atmosphere saturated with aqueous 
vapour, which the natural atmosphere almost never is. In order to 
obtain such an atmosphere, I proceeded in the usual manner : — A glass 
cylinder, .the lower half of which was lined with moistened blotting 
paper, was inverted over the hygrometer. The index began to move, 
and continued to do so for a while, but then stopped, although it was 
evident that the contained air was far from being saturated. , I soon 
found that this discontinuance of the effect of the moistened blotting 
paper was dependent on the nature of the substance on which the glass 
cylinder was inverted. If on wood, the interruption took place ; if on 
glass, the index continued to turn until the elasticity of the gut-line 
was exhausted. The cause of the difference is obvious : wood is an 
absorbent of moisture. 
In experimenting with an artificial atmosphere the effect on the 
secondary index was not, in all cases, such as to produce equal diffe- 
rences. On five rounds of the primary index the differences shown by 
the secondary were the following : — 
Primary completed 100° Secondary 10° Difference 10° 
23 „ 13 
40 18 
60 ,, 20 
82 „ 22 
This irregularity, I at length perceived, was dependent on the 
part of the glass cylinder occupied by the moistened blotting paper. 
200 
300 
400 
500 
