208 Proceedings of the Royal Irish Academy. 
festly such a state of tlie atmo spheric air introduced into the lungs as 
would be competent by its greater weight and expansile capacity, to 
increase the pressure upon their internal surface, would constitute a 
cause of this kind, and be exemplified in the breathing of com- 
pressed air. 
It seems clear, therefore, that the lungs are the organs primarily 
and most directly affected by the inhalation of compressed air, and that, 
curatively, any condition of the lungs involving impairment of their 
capacity for elastic reaction, or permanent expansion of their structure, 
would be unsuited for the therapeutic application of compressed air. 
Yet I find Dr. Burden- Sanderson declaring his opinion that it would 
be useful in the treatment of pulmonary emphysema, and chronic bron- 
chitis, in which there is morbid and permanent expansion of the 
lungs owing to impairment or loss of elasticity of their structures, 
and in which, moreover, the right chambers of the heart and the 
systemic veins are engorged with blood. I need scarcely say that I 
cannot subscribe to this opinion. I believe that in such a condition of 
the lungs and right heart, the respiration of compressed air would 
greatly aggravate the evil proposed to be remedied. I think it likely, 
however, that if a suitable appliance were devised, by which com- 
pressed air, exceeding a pressure of two and a half atmospheres, or of 
still greater density, according to the object sought to be attained, 
could be introduced into the lungs, whilst the body was subjected to 
ordinary atmospheric pressure, great benefit would arise therefrom in 
the treatment of atelectasis, or chronic tuberculosis of the lungs, or 
of pleuritic effusions of long standing with collapse or compression of 
the lung. 
XXXII Oif A ISTew Poem of Spectroscope. Ey G. Johnston^e 
Stoi^ey, M. a., T. R. S. 
[Eead June 26, 1871.] 
1 . If ^ be the minimum deviation of a ray in a prism, or a battery 
of prisms, and i its inverse wave-length, i. e. the reciprocal of its wave- 
length, then 
may be taken as a convenient measure of the dispersion. 
This will -assume a numerical form, if we measure the deviations 
in tenths of a minute of arc, and the wave-lengths in fractions of 
a millimetre. The inverse wave-lengths will then be the denomina- 
tors of these fractions, and h will be the number of tenths of a minute 
of arc, over which one unit of this scale of inverse wave-lengths is 
dispersed. 
The value of b so defined will vary from point to point along the 
spectrum, having usually between two and three times the value 
