Art. XI.] Cole, Experience with Capillary Electrometer. 269 
In observing the deflections a microscope with eyepiece 
micrometer was used, i960 of whose divisions on the rotating 
head are equal to i mm. 
Thus the hope of securing high sensitiveness with this 
form was realized. Tube i was of nearly the same size as that 
used by Paschen, while tube 6 with a capillary of about one- 
fifth millimeter, showed a sensitiveness more than 40 times as 
great. (It should be mentioned in this connection that sensi- 
tiveness was not an object in Paschen’s research, but it was nec- 
essary to have the capillary of large size in order to allow the 
gas formed by the high voltages used, to escape.) 
The most surprising feature of the above table is the indi- 
cation of maximum sensitiveness for a capillary of about one- 
fifth millimeter diameter. The extreme sensitiveness of tube 
No. 6 is however probably due in part to its form, for another 
of nearly the same diameter but of somewhat different taper 
gav^e a deflection only one-third as great. A number of ob- 
servers have called attention to the influence of the shape of 
the capillary tip in instruments of Ostwalds third form, and 
many of my own measurements show it. 
No. 6 not only shows great sensitiveness, but (in sharp 
contrast with those of similar delicacy of the ordinary type) 
does not easily get out of order and can be taken apart, washed 
and dried, and show its former sensitiveness when set up anew. 
To test its durability and constancy under such conditions, the 
following course was adopted. After its sensitiveness had been 
measured, it was short circuited, and the next day its constant 
was redetermined. It was found unchanged. Another wait of 
two days showed its sensitiveness slightly increased. Several 
weeks later it was taken apart, washed and shipped to a neigh- 
boring state. After eight months it was again set up with dif- 
ferent mercury and acid, and the sensitiveness when measured 
was found to be nearly the same as before. 
The resemblance of this type to Ostwald’s second form is 
rather close. The shorter length of its capillary is a consider- 
able advantage however, provided the capillary is fine. Break- 
ing or clogging of the mercury column is less likely to occur 
