1877.] 109 [Chase. 
burg’s extremes (hydrogen .57, air .76) seem to point towards secondary 
nucleal and atmospheric relations between air and hydrogen. 
In my identification of the velocity of solar dissociation with the velocity 
of light, * although the conception of successive wave impulses seems most 
natural, itis by no means essential. If the pressure of the ultimate force is 
constant, the result is the same. Tlie ratio of the velocity of dissocia- 
tion to the velocity of perfect fluidity, is approximately illustrated by 
Draper’s estimate of the ratio between the temperature of glow (977° F., 
or 1436° from absolute 0°) and the temperature of fluidity (82° F., or 
491° from absolute 0°; 1436 ~ 491 = 2.9). Here complete fluidity is 
compared with incipient glow. The ratio 7: 1 would require an addi- 
tional allowance of 107°, or about 7.5 per cent., for the difference between 
the temperature of complete and incipient glow. Ifthe comparison were 
made at 0° F., we should have 1436 — 459 = 3.13. 
The vis viva of terrestrial dissociation being equivalent to 4 the v. »v. 
of incipient planetary dissociation at the Sun,{ the temperature ratio of 
water vaporization to dissociation furnishes another illustration of a simi- 
lar character. Deville (C. Rendus, lxxxiv, 1259) quotes the estimates made 
by himself and Debray (2500°), and by Bunsen (2800°), of the temperature 
at which nearly half of the vapor of water is reduced to its elements, hydro- 
gen and oxygen. The ratio 2800° : 100° is a very probable estimate of the 
ratio between solar and terrestrial superficial gravitation. 
Notre.—August 23, 1877. In consequence of a remark near the opening 
of the foregoing paper, Dr. Draper recently proposed that I should test 
some of my views by an examination of the solar spectrum. I accord- 
ingly undertook a preliminary investigation, which has already yielded the 
following results : 
; : c C c 
In the harmonic progression, —> , 
n n+a n+2a 
length of Fraunhofer line A = 761.20 millionths of a millimetre ; » — 
1.0153 ; a = .0918 ; and we find the following accordances: 
» €tc., let c = wave- 
Numerator. Denominators. Quotients. Observed values. 
761.20 n+ a 687.75 687.49 — B 
nt 3a 589.89 589.74 = D1 
n+ 6a 486.14 486.52 = 
n+ 104 393.79 393.00 = Ht 
The ‘‘observed values’’ are the wave-lengths, as determined by Dr. 
Wolcott Gibbs (Amer. Jour. Sci. [2] xliii, 4). The lines between A and 
B have not been studied sufficiently to fix their wave-lengths ; it seems 
likely that A ~- may bea bright line, and thus belong to the field of 
investigation which Professor Draper has so brilliantly opened. The 
greatest difference between the above theoretical and observed values, is 
* Ante, xvi, 307, et al. + Ib. p. 305. { Ib. p. 307-8. 
