982 8. W. Holman—Method for Calibrating Thermometers. 
termediate points: in other words, to shift the axis of the 
second curve of error so that it shall make the error at B zero. 
his method requires for each calibration the use of but a 
single thread. The computation is simple, and involves a mini- 
mum of approximation. Hrrors of observation are largely 
eliminated by the number of settings made in all parts of the 
tube, and by the inspection of the curve of lengths, both of 
ing simple met ; 
Considerable aid in eliminating errors of parallax in such 
work is sometimes found by looking down upon the horizontal 
thermometer through a vertical tube having a small hole at 
each end. One of the cheap French microscopes with its 
lenses removed, and inverted in its stand, answers this purpose 
well. 
With such a device two calibrations of the above described 
thermometer with threads of 3 cm. and 5 cm. respectively, each 
with only one series of observations, and requiring not more 
than one hour and a half each for completion, gave results 
whose average difference from each other at nine points was 
0:04 mm., and the arithmetical sum of the extreme differences 
was 012 mm., a result of sufficient accuracy for any class of 
work of which such an instrument is capable. 
For brief descriptions of methods of separating threads of 
mercury for calibration, reference may be made to the paper 
