THE VISCOSITY OF WATER. 35 
'6°494, 5°408, and 6°456 cms. respectively at 0° C., and their 
radii approximately were °019, °019, °020, °020 cms. respec- 
tively, at 0°C. (The exact determination of the equivalent 
radius of each capillary was made at the end of all the 
‘experiments. It involved the cutting up of the tube, and 
the careful measurement of the sections). Sufficient 
freshly distilled water was put in to fill up the glischro- 
‘meter from b to c. 
The ends A and B were connected to the reservoir of 
compressed air or the outside air by means of three-way 
taps. Measurements of viscosity were taken first with the 
water flowing out of R into L, and secondly with the 
water flowing in the opposite way, under the pressure 
of air in the reservoir. The average of the two deter- 
minations was taken as an absolute value of the viscosity. 
The pressure was measured by means of (1) a water mano- 
meter 200 cms. long, (2) a mercury manometer in cases 
where the pressure was greater than 200 cms. of water. 
To facilitate the reading of the water manometer, a pair 
of small parallel mirrors was attached to each of the arms. 
These were inclined so as to make an angle of 45° with the 
vertical manometer scale. One was fixed near the centre 
of the arm, and opposite a telescope; the other could be 
moved along the arm and clamped in front of the water 
surface. When the mirrors had been set, it was thus 
possible to read off the positions of the two water surfaces 
by means of the telescope, at the same instant; for both 
images were arranged to be side by side in the field. The 
pressures employed varied between the limits 100 cm. 
(water) and 42 cms. (mercury). 
The time of flow was in most cases very short, the aver- 
age being about one minute, but in extreme cases it was 
as low as 22 seconds. Special means had to be employed 
to register the time intervals correctly. The chronograph 
