210 
LOl;i) IlAYLEIGII OX A XEA' MAXOMETEE, 
the centres of vrhicli are situated tlie points. These are of glassA" which need 
not he opatjiie ; and tliev must he carefully finished upon a stone. A considerable 
degree of sharpness is desirable, but siniilcivitij is more important than the extreme 
of sharpness. In the actual apparatus complete simila,rity was not attained, and 
in the first trials the ditference was rather emharrassim;’. However, after a little 
})ractice the eye becomes educated to set tiie mercury to each point in a constant 
manner, and this is all tliat is really required. The same consideration shows 
that minute outsta.iiding capillary differences sliould not lead to error. It may 
be remarked that the mercury is always on the rise at the time of adjustment, 
and in fact it was found best to make it a rule not to allow the points to be 
drowned at anv time when it could be avoided. After such' a drownino- it was* 
V O 
usually (perha})S alvaiys) found that the mercury surface was disturbed by the 
proxunitu of the points without actual contact, an effect attrihuted to electrification. 
The presentation of the jioint to the mercury, or rather of the point to its image 
as seen by reflection in the mercury, was examined with the aid of two similar eye- 
lenses (not showii) of 22 millims. focus. The illumination, from a small gas flame 
^ At first iron iicuiUo points were trieil. 
