452 
Proceedings of the Royal Society of Edinburgh. [Sess. 
Again, if V be the velocity in miles per hour, we have 
Y = 60a sin 0/(tf 3 - £ 2 ) ; 
that is, 
Y = 336 sin 0/(t 3 - t 2 ) (42). 
On the assumptions we have made, the formulae (41) and (42) enable us 
to calculate the direction and velocity of propagation of any disturbance, 
the same phase of which has been observed at A, B, and C. 
It may be of some general interest, and possibly also of some use to 
others who may investigate the minor fluctuations of the atmospheric 
pressure with more perfect means in the future, to give a brief account of 
the Earn observations, and to discuss how far it is possible to deduce 
definite conclusions from them. 
In the first place, it should be mentioned that, of the three instruments 
used, two came into the possession of the observers of the Lake Survey 
after they had started work on Loch Earn. There was therefore no 
opportunity of testing the three instruments together in the laboratory 
beforehand as regards sensitiveness, equality of leak, regularity of 
clock, etc. 
The time-scale of these instruments is unfortunately very small, viz. 
•25 mill, per minute ; so that it is barely possible to make determinations 
of time to a minute on the microbaro^ram. It should also be mentioned 
that, since we had no direct telegraphic or telephonic communication 
between our three observing stations, it was difficult to get frequent and 
reliable comparisons of the watches of the observers. On the ground of 
the theoretical calculations in the first part of this paper, and of laboratory 
comparisons of the three instruments which have recently been made, it 
may be assumed that the differences of the acceleration of phase in the 
three microbarograms, whether due to the differences of leakage or to 
progressive deformation of the disturbance (which undoubtedly takes place 
in many cases), amounted to less than one minute. Also, our watch 
comparisons, although they left much to be desired, are accurate to less 
than a minute. Unfortunately, the clockwork supplied with the instru- 
ments did not work in all cases with sufficient regularity, and this, 
notwithstanding the use of frequent time-marks, introduced a considerable 
amount of uncertainty into our results. 
On account of the necessity of frequent time-comparisons, and also for 
the purpose of calling the simultaneous attention of observers to any 
phenomenon of interest the beginning of which may have been noticed at 
any station, I would strongly advise that any network of microbaric 
stations should be connected by telephone. 
