1907-8.] On the Theory of the Leaking Microbarograph. 437 
XXX— On the Theory of the Leaking Microbarograph; and on 
some Observations made with a Triad of Dines - Shaw 
Instruments. By Professor Chrystal. 
(MS. received June 8, 1908. Read June 15, 1908.) 
The minor fluctuations of the atmospheric pressure are now engaging 
the close attention of meteorologists, and my own attention has been 
drawn to them on account of their connection with the oscillations of 
lake-surfaces generally known since Forel’s investigations under the name 
of seiches. 
As these fluctuations often do not exceed a millimetre or two of water, 
they are not shown by the ordinary self -registering apparatus. In order 
to record them we may use a specially sensitive form of barograph, such 
as the Richard statoscope, which is delicate enough to show these small 
fluctuations, and yet can be brought back to a momentary zero whenever 
the indicator threatens to go off the scale or beyond the limits of safety. 
There are, however, two objections to this method of measurement. In 
the first place, the instrument must either be watched or else provided 
with a self-acting arrangement for altering the zero. But there is a much 
more serious and radical objection, which will at once be understood by 
anyone who has studied a tidal curve showing what are called “ secondary 
tidal oscillations,” i.e. certain oscillations of much smaller range than the 
tidal oscillation proper, and of much shorter period, say fifteen to twenty 
minutes. Fig. 1 is a reproduction of such a curve. At first sight it 
Fig i 
would appear that the secondary oscillations occur only near high and 
low water. As a matter of fact, however, they occur throughout the 
whole day, and are merely masked in the mareogram at the zero of the 
tide, because the rapid rise of the curve due to the proper tide wipes out 
the turning-points and even the inflexions due to the minor oscillation. 
It is necessary, therefore, to devise an instrument which will neglect the 
