PHYSIOLOGY: P. D. LAMSON 
521 
a'^du^^ la (H - K) du dv^-'^dv'^^O (15) 
and defines the pair of lines through y which separate harmonically 
both the pair of anti-ray tangents and the pair of associate conjugate 
tangents of the point y. 
The two Jacobians (14) and (15) coincide if and only if 
~^loga = 0, (3 to) 
i.e., if and only if the original conjugate net is iso thermally conjugate. 
We may state our result as follows: 
A necessary and sufficient condition that a conjugate net of curves on a 
surface he isothermally conjugate is that at each point of the surface the pair 
of axis tangents, the pair of associate conjugate tangents, and the pair of 
anti-ray tangents he pairs of the same involution. 
By means of the various nets of curves defined in the course of the 
above interpretation, we have been enabled to deduce a number of 
properties of isothermally conjugate nets. We have included this more 
extended discussion in a longer paper, which is a sequel to the one on 
conjugate nets to which reference has already been made. 
^ L. Bianchi, Vorlesungen iiher Diferentialgeometrie, tr. M. Lukat, 2te Aufl., pp. 135 
et seq. 
2E. J. Wllczynski, Trans. Amer. Math. Soc, 16, 311-327 (1915). 
3G. M. Green, Amer. J. Math., 37, 215-246 (1915). Cf. §1. 
^Ibid., end of §3. 
THE ROLE OF THE LIVER IN ACUTE POLYCYTHAEMIA : THE 
MECHANISM CONTROLLING THE RED CORPUSCLE 
CONTENT OF THE BLOOD 
By Paul D. Lamson 
PHARMACOLOGICAL LABORATORY. JOHNS HOPKINS UNIVERSITY 
Presented to the Academy, August 18, 1915 
It is very generally considered by all except those who have paid 
special attention to the subject, that the number of red corpuscles per 
unit volume of blood is, in the normal individual, a fairly fixed quan- 
tity subject to gradual change only. A more careful study shows how- 
ever that this number is subject to very rapid and great changes, and 
instead of being constant, that it is continually changing under physio- 
logical conditions. 
Questions naturally arise as to what factors will cause a change in 
