148 Proceedings of Royal Society of Edinburgh. [sess. 
bore has a capacity of c.c., the other c.c. The others of the 
set have capacities of 1 c.c. and 5 c.c. 
There is a pipe, not shown in sketch, connecting A to A, outside 
the stopcock. This pipe is of smaller bore than A, and its object 
is to allow a constant current of air to be kept flowing through the 
pipe while the measuring stopcock is closed in that direction. This 
pipe is most easily connected a little on each side of AA. 
As to the relative advantages of the two methods of measuring 
the air, the cylinder plan has the advantage that the one apparatus 
measures all quantities without any change in the arrangements ; and 
further, it does not confine us to certain fixed quantities. It also 
enables us to check our results, by using twice or any suitable mul- 
tiple of the dusty air, and for experimental purposes it seems to be 
the best. On the other hand, the stopcock plan seems to be the 
most accurate, as, practically, no time is given for the dust to settle 
in the measure. If the air to be tested is very impure, so that 
the quantities to be measured are very small, such as c.c. and 
under, and if the impurity of the air does not vary much, then the 
stopcock measure is the best. It is also much more easily worked 
than the other, fewer movements being required for each test. On 
the other hand, if the quantity of dust varies much, it necessitates 
a change of measure to meet the changing conditions, a large stopcock 
being necessary when the air is pure, and a small one when dusty. 
In gauging the bore of these stopcocks, for the large sizes I have 
generally used measurement and checking by gauging with water. 
The small ones have been gauged by means of wires. The bore was 
first drilled a little smaller than it was ultimately to be. From this 
bore the length was got which enabled the proper diameter to be de- 
termined. The following plan of determining the diameter was found 
of easy application : — Take the case of the stopcock sketched in 
Plate III. The length of the larger of the two bores was 2 cm., and 
it was required to have a capacity of i c.c. From a bundle of steel 
wires of different sizes one was selected, the volume of which was 
1 c.c. per 10 cm. of its length. This was ascertained by selecting a 
wire which, when dipped into a burette nearly full of water, displaced 
1 c.c. per 10 cm. of its length. The bore in the plug of the stopcock 
was then made to fit this wire exactly, and, as the bore is 2 cm. 
long, its capacity will be } c.c. 
