438 
Proceedings of the Koyal Society of Edinburgh, [Sess. 
barometric variations of larger range and longer period, and record only 
the minor fluctuations. This is effected by a method which is, in a sense, 
the converse of that which I described in my memoir on the Investigation 
of the Seiches of Loch Earn * for damping out the seiches of higher 
nodality in a limnogram. 
This method consists essentially in measuring the difference between 
the atmospheric pressure and the pressure in a vessel which communicates 
with the atmosphere by means of a small leak, say through a capillary 
tube of sufficient length or fineness of bore. If the bore were infinitely 
fine, the instrument would simply register the atmospheric pressure with 
all its variations ; if the bore were very wide, it would register nothing at 
all ; and by properly adjusting the tube we can arrange so that it only 
registers the fluctuations of short duration, which pass away before the 
small flow through the capillary has had time to establish equality of 
pressure within and without. 
Theory of the Dines-Shaw Microbarograph. 
The instrument, which is represented diagrammatically in fig. 2, 
consists essentially j- of an air-chamber V of considerable capacity (about 
*3 cub. ft. = 9440 c.c.), carefully packed to reduce variation of temperature 
as much as possible. This communicates through a narrow pipe PP with 
the small air-space under an inverted cylindrical cup, which floats in 
mercury contained in a cylindrical cup A A. The vertical motions of the 
cup CC caused b}^ the variations of the pressure outside and inside V are 
registered by means of a magnifying lever LL, into the details of the 
construction of which it is unnecessary to enter here. 
The air-chamber V communicates with the outer air by means of a 
capillary tube KK, through which the air leaks slowly outwards or 
inwards according as the pressure outside is less or greater than the 
pressure inside. 
We shall suppose that the air-space under the floating cup CC is 
practically negligible in comparison with V, and also that the temperature 
is constant. 
We shall also suppose that the leak is so slow that the cup may be 
supposed to adjust itself instantaneously to the pressure conditions at any 
moment ; and we shall also neglect the inertia and friction of the movable 
parts of the apparatus. 
* Trans. Roy. Soc. Edin vol. xliv. p. 373 (1906). 
t See Quart. Journ. Roy. Met. Soc., vol. xxxi. p. 39 (1905). 
