May 6, 1915] 
NATURE 
263 


which raises the pen by a step at a time, the value 
of the interval shown by the step being the capacity 
of the bucket, which empties as it tips. The best 
instrument of this type is that in use by the United 
States Weather Bureau, which has a bucket tipping 
with one-thousandth of an inch of rain, and so 
gives a fairly continuous line, even in moderately 
light showers. Attractive and well-made tipping- 
bucket rain gauges have been put on the market 
by the firms of Negretti and Zambra, and Pastor- 
elli and Rapkin, of London, and by Richard 
Fréres in Paris. These tip with one-hundredth of 
an inch, or sometimes with half that amount, but 
are useless for measuring duration of any but ~ 
heavy rain, though under careful inspection fairly 
satisfactory for measuring amounts. The amount 
of rain can always be measured more accurately 
by means of a direct reading rain gauge of the 
Snowdon or Meteorological Office pattern, of 
which the former is, in my opinion, better as well 
as cheaper. 
The only scientific purpose served by a recording 
rain gauge is 
to furnish a 
measure of dura- 
tion and inten- 
sity. The two 
types of gauge 
involving the 
respective use of 
a counterpoised 
receiver or of a 
float give their 
record in the 
form of a con- 
tinuous curve. 
Each type has 
several modi- 
fications ren- 
dered necessary 
by the practical 
convenience — of 
using a_ shallow 
drum while 
retaining an 
‘open scale. Where expense and space are no 
objects rain gauges of unimpeachable accuracy 
can be made on either principle by having a drum 
deep enough to record the whole rainfall of the 
wettest day possible with the necessary degree of 
magnification. 
In order to keep the moving parts compact 
and the drum small, the capacity of the receiver 
is in practice usually limited to from 020 in. in 
the case of a counterpoised vessel to 0°50 in., or 
rarely 1 in., in the case of a float, and an auto- 
matic contrivance for emptying the vessel when- 
ever it fills has to be employed. This is the weak 
point of most recording rain gauges, for when 
the receiver has to be emptied five times for every 
inch of rain it is subject to much friction and wear, 
and when it empties only once or twice for an inch 
of rain the time of clearing is appreciable and 
introduces a risk of error. 
The Beckley rain gauge made by Mr. Hicks 
NO. 2375, VOL. 95] ‘ 
























Fic. 1.—New Casella recording rain-gauge, 


and used in the observatories of the Meteorological 
| Office for about forty years is the best example 
of a counterpoised receiver writing directly on the 
drum as it sinks vertically and discharging auto- 
matically by means of a self-starting syphon. The 
‘ Casella recording gauge, which has been in use 
at Camden Square for about thirty-five years, is 
| undoubtedly the best example of the counterpoised 
| receiver emptying by automatic tipping when it 
comes to the lowest point. It writes by means of 
| a system of levers on a drum with a horizontal 
gauge (Fig. 1), 
axis. A modified form of this 






























6 
Necrett! & ZAMBRA LONDON.. 

3-—Mechanism of the Fernley 
recording rain-gauge. 
3 2.—Section of Halliwell standard Fic. 
recording rain-gauge. 
| recently introduced by Messrs. Casella at a low 
| price, has a simpler mechanism, the receiver only 
tipping far enough to start a syphon, through 
which it is emptied, while the pen writes on a 
vertical drum by means of a hinged piece at right- 
angles to the main lever, thus securing a curve 
with approximately rectangular co-ordinates. 
Of float rain gauges the best of which I have had 
experience is the improved form of the Halliwell 
rain gauge (Fig. 2) first constructed by Mr. F. L. 
Halliwell with the advice of Mr. Baxendell, the 
well-known director of the Fernley Observatory, 


