] 904 - 5 .] Dr J. Halm on Astronomical Seeing. 461 
The worst definition was commonly experienced shortly before 
sunrise. Professor Copeland tells me that at Parsenstown the 
seeing was specially good during a gale, and my own experience 
here in Edinburgh confirms this statement. 
The superiority of the definition in summer over that in winter 
which is very marked at continental observatories is readily 
explained by the fact that convection is much more energetic in 
the former season. Indeed, at continental stations the atmo- 
sphere in winter is on the whole very far from the condition of 
adiabatic equilibrium, the temperature-gradient being much too 
small, and often even reversed. 
The question is doubtless of practical importance, and should 
receive attention when sites for new observatories are selected. 
The erection of observatories on or near mountains may be 
advocated from this point of view, because horizontal movements 
of the atmosphere are deflected at the mountain sides into more 
vertical directions, thus enhancing that “stirring” of the atmo- 
sphere above the station which leads to the establishment of con- 
vective equilibrium. The atmosphere on mountains, besides being 
more transparent, must also be steadier, in an optical sense, not 
from the absence of motions, but because these motions, by taking 
place under adiabatic conditions, exert little or no disturbing 
influence on the normal refrangibility of the air. 
Meteorologists may perhaps give us definite and practical hints 
as to the more or less favourable conditions under which convection 
takes place in our atmosphere. Astronomers should be guided by 
these advices in the selection of localities for their observatories. 
Clearly, we have no means to prevent the incessant general and 
local movements of the vast gaseous ocean above us. But knowing 
that under one certain condition these uncontrollable motions, 
otherwise so much inclined to impair our vision, may be rendered 
optically ineffective, we must avail ourselves of every possible 
chance by which this ideal condition may be approached, — on the 
one hand, by taking full advantage of favourable topographic and 
climatic features, and on the other, by designing mechanical devices 
for inducing convection in the neighbourhood of our instruments. 
It would be interesting to hear the opinion of practical 
astronomers on this question, and to see how far their experiences 
