4i 6 Baker . — Quantitative Experiments on 
the other is through the long bent capillary tube. By means of a well-known 
physical formula 1 it is possible to calculate the length and bore of tubing 
required to allow any given volume of air to enter at any temperature under 
the driving pressure represented by the height of the column of water in 
the reservoir. The necessary length is rather unwieldy for the small 
amounts of water (about io grammes per fortnight) required in these 
experiments. It works out to about 50 centimetres of 0-05 mm. bore 
capillary. But as the air admitted varies as the fourth power of the 
radius, and inversely as the length, a small decrease in the bore of the 
capillary is equivalent to a large increase in length, and so a convenient 
length of 0*5 mm. tubing was used, and this was then drawn out into 
exceedingly fine capillary at the end. The adjustment is made by breaking 
off small pieces of the fine capillary. Then the volume of water expelled is 
exactly equal to the volume of air admitted. The apparatus works admir- 
ably until the capillary tubing becomes blocked by condensed water, and 
the small bulb leading into the capillary tubing was filled with glass wool to 
prevent this as long as possible. No doubt this principle could be applied 
to a larger watering apparatus, the air being admitted through capillary 
tubing of suitable length, waxed into a well-fitting cork. 
In actual practice it was found, however, that the other entry for air, 
through the solution itself, allowed sufficient water to escape, without the 
adjustment of the capillary, when the temperature underwent the usual 
diurnal variations, and so the end of the capillary was sealed over. The air 
above the water in the reservoir became so rarefied in the cool of evening 
that air bubbles were sucked up through the liquid, with the result that 
next day, as the air warmed again, water was expelled. This method of 
watering was the more convenient because the water-supply was automati- 
cally regulated to the needs of the plants ; on hot days much water was 
expelled, on cool days little. But it could not be used to give much more 
water than about 15 grammes a week in summer unless a huge air-bulb 
were made at one side, and, of course, in a carefully regulated laboratory, 
where the temperature could be kept fairly constant, it would prove 
inadequate even for the needs of mustard seeds. 
As a rule, one waterer was set above each culture, one culture occupying 
each bell-jar, but in one or two of the final experiments it was found con- 
venient to make large waterers — the reservoirs about 4 cm. in diameter 
and 15 cm. high — which would hold sufficient water to last four or five 
weeks for two cultures, and have two cultures in each bell-jar. The water 
1 The volume of gas F y passing per second through a tube of length l and radius r , under a 
difference of pressure {p x — p 2 ) is represented approximately by the equation p 2 V = 
(A 2 -A 2 ) 
/j^ 2 ) 1*3 * V. p UV.U. 1 J J & A ' J, 7 
where rj is the coefficient of viscosity of the gas. (See Poynting and Thomson : Properties of Matter, 
p. 212, London, 1905.) 
