VISCOSITY OF WATER BY THE EFFLUX METHOD. 145 
rapid to allow of a record to a small fraction of a second,' whereas 
the total time of efflux is more than 8000 seconds or over two 
hours. The correction of the efflux time for fall in pressure at 
the entrance of the tube is about 24900 sec./7’, or about 1/2700th 
for 0° C.—formula (5)—or the correction to be applied to the 
deduced viscosity about 0:0445/7'—formula (40)—-so that a small 
uncertainty as to its proper magnitude will not greatly vitiate 
the result. 
The volume of the bulb V would be ascertained readily and 
with sufficient accuracy from its capacity, by using quicksilver. 
The measurement on the accuracy of which the result mainly 
depends is the capillary radius, 0-1 millimetre.. To obtain its 
value to considerably under 1/10000 th part? is not an easy task. 
The whole volume of the tube represents only 0-214 gramme of 
mercury, so that on this element of the work—zi.e, the evaluation 
of the ‘mean radius of efflux’ of the tube—by far the greatest 
expenditure of effort to attain precision will be required. This 
applies when the absolute determination of the viscosity is 
required 
The following series of experiments seems to me to meet the 
requirements of a thorough determination : 
Temp. every 5°, 10°— 95°C.; H = 100; LZ = 50 
every 0°.5,0°-10°C. H= 50; L = 50 
and H = 100; ZL = 25 
also at 4°C. H = 100, 50, and 25; Z = 50 
and H = 100 LZ = 50, 25 and 125 
Such @ series ought, with care as to dimensions, careful ther- 
mometry and generally good manipulation, to satisfactorily settle 
. MAREE SOR 
‘1The supply tube may be of any convenient size, but it is always 
“mad that at the points I and J the velocity shall be sensible when 
ed by the microscope R. If constricted great care should be taken 
that then nstricted diameters are equal so that the concavity of the water 
may be the same at each point. 
y ding to a precision of over 1/2500 th. 
J—Iuly 3, 1895, 
