So far, it appears only that we have been measuring heat, but I have called the curve that of solar "energy," because by a series of independent investigations, not here given, the selective absorption of the silver, the speculum-metal, the glass, and the lamp-black (the latter used on the bolometer-strip), forming the agents of investigation, has been separately allowed for. My study of lamp-black absorption, I should add in qualification, is not quite complete. I have found it quite transparent to certain infra-red rays, and it is very possible that there may be some faint radiations yet to be discovered even below those here indicated.

In view of the increased attention that is doubtless soon to be given to this most interesting but strangely neglected region, and which by photography and other methods is certain to be fully mapped hereafter, I can but consider this present work less as a survey than as a sketch of this great new field, and it is as such only that I here present it.

All that has preceded is subordinate to the main research, on which I have occupied the past two years at Alleghany, in comparing the spectra of the sun at high and low altitudes, but which I must here touch upon briefly. By the generosity of a friend of the Alleghany Observatory, and by the aid of Gen. Hazen, Chief Signal Officer of the U S. Army, I was enabled last year to organize an expedition to Mount Whitney in South California, where the most important of these latter observations were repeated at an altitude of 13,000 feet. Upon my return I made a special investigation upon the selective absorption of the sun's atmosphere, with results which I can now only allude to.

By such observations, but by methods too elaborate for present description, we can pass from the curve of energy actually observed to that which would be seen if the observer were stationed wholly above the earth's atmosphere, and freed from the effect of its absorption.

The salient and remarkable result is the growth of the blue end of the spectrum, and I would remark that, while it has been long known from the researches of Lockyer, Crova, and others that certain rays of short wave-length were more absorbed than those of long, these charts show how much separate each ray of the spectrum has grown, and bring, what seems to me, conclusive evidence of the shifting of the point of maximum energy without the atmosphere toward the blue. Contrary to the accepted belief, it appears here also that the absorption on the whole grows less and less, to the extreme infra-red extremity; and on the other hand, that the energy before absorption was so enormously greater in the blue and violet, that the sun must have a decidedly bluish tint to the naked eye, if we could rise above the earth's atmosphere to view it.

But even were we placed outside the earth's atmosphere, that surrounding the sun itself would still remain, and exert absorption. By special methods, not here detailed, we have at Alleghany compared the absorption, at various depths, of the sun's own atmosphere for each spectral ray, and are hence enabled to show, with approximate truth, I think for the first time, the original distribution of energy throughout the visible and invisible spectrum at the fount of that energy, in the sun itself. There is a surprising similarity, you will notice, in the character of the solar and telluric absorptions, and one which we could hardly have anticipated a priori.

Here, too, violet has been absorbed enormously more than the green, and the green than the red, and so on, the difference being so great, that if we were to calculate the thickness of the solar atmosphere on the hypothesis of a uniform transmission, we should obtain a very thick atmosphere from the rate of absorption in the infra-red alone, and a very thin one from that in the violet alone.

But the main result seems to be still this, that as we have seen in the earth's atmosphere, so we see in the sun's, an enormous and progressive increase of the energy toward the shorter wave-lengths. This conclusion, which, I may be permitted to remark, I anticipated in a communication published in the Comptes Rendus of the Institute of France as long since as 1875, is now fully confirmed, and I may mention that it is so also by direct photometric methods, not here given.

If, then, we ask how the solar photosphere would appear to the eye, could we see it without absorption, these figures appear to show conclusively that it would be blue. Not to rely on any assumption, however, we have, by various methods at Allegheny, reproduced this color.

Thus (to indicate roughly the principle used), taking three Maxwell's disks, a red, green, and blue, so as to reproduce white, we note the three corresponding ordinates at the earth's surface spectrum, and, comparing these with the same ordinates in the curve giving the energy at the solar surface, we rearrange the disks, so as to give the proportion of red, green, and blue which would be seen there, and obtain by their revolution a tint which must approximately represent that at the photosphere, and which is most similar to that of a blue near Fraunhofer's "F."

The conclusion, then, is that, while all radiations emanate from the solar surface, including red and infra-red, in greater degree than we receive them, the blue end is so enormously greater in proportion that the proper color of the sun, as seen at the photosphere is blue--not only "bluish," but positively and distinctly blue; a statement which I have not ventured to make from any conjecture, or on any less cause than on the sole ground of long continued experiments, which, commenced some seven years since, have within the past two years irresistibly tended to the present conclusion.

The mass of observations on which it rests must be reserved for more detailed publication elsewhere. At present, I can only thank the association for the courtesy which has given me the much prized opportunity of laying before them this indication of methods and results.