probability may be inferred that, with a longer ex-* posure, the rays would have been presented as seen by Gilman, Eastman, and others, if not as seen by General Myer. In the photograph there is, indeed, a sharpness of outline which might readily be interpreted by those unacquainted with the nature of photographic processes to imply the existence of a real boundary line separating this part of the corona from the part without. But, as a matter of fact, the sharpness of outline is due to peculiarities in the process of development. It may be recognised in the photographs taken at Ottumwa, although in them the corona has a much smaller extent. It is not noticed, however, in the photographs by Dr. Curtis, where the corona has about the same degree of extension; the reason being, that he employed special care in avoiding over-development of the negative. Hence no doubt whatever can remain that the sharpness of outline in the Ottumwa photographs, as also in Mr. Whipple's, implies no real limitation of the object photographed.*

* As some stress has been laid on this matter by those who advocate theories respecting the terrestrial nature of the corona, it may be well to present at length Dr. Curtis's statement respecting the erroneous interpretation of these photographic records. He says that he has read with surprise an extract from a letter, written by Dr. Gould to Professor Henry Morton, in which the former says, An examination of the beautiful photographs made at Burlington and Ottumwa, by the sections of your party in charge of Professors Mayer and Himes, and a comparison of them with my sketches of the corona, have led me to the conviction that the radiance around the Moon, in the pictures made during the totality, is not the corona at all, but is actually the image of what Mr. Lockyer has called the chromosphere.' 'Dr. Gould proceeds,' says Dr. Curtis, to specify the points at variance between the corona as photo

Such is a sketch of the evidence adduced up to the present time respecting the solar corona. It appears

graphed and the same object as seen and sketched by him; and because the two representations do not correspond in feature, he infers that the objects depicted cannot be identical. This same argument would apply equally well to the "radiance" shown in my own photographs, since in them the phenomenon, though faint, agrees in outline with the [similar object on the Burlington and Ottumwa pictures. Now, I cannot but believe that Dr. Gould is in error in imagining this aureola not to be simply the image of the more intense portion of the corona near the surface of the Sun. In the first place, the experience of this very eclipse has shown how guardedly all sketches and drawings of the appearances of totality should be received, as affording an accurate record of either the shape, size, or position of the various objects. This is evident upon comparing the various sketches made by eyeobservers of the protuberances and corona, both with each other and with the photographs, and observing the very great discrepancies manifest. Of course, it is not meant that accurate measurements made by a micrometer eye-piece in the telescope, or similar determinations of position-angle, cannot be relied upon, but, on the contrary, the argument is that only such are to be received as trustworthy, and that all general sketches and drawings made hastily during the few exciting minutes of totality, or from memory afterwards, form but a weak ground upon which to base an important scientific hypothesis. But positive proof in the question at issue is afforded by the very perfect photographs of the corona taken at Shelbyville, Kentucky, by Mr. Whipple, of the Cambridge expedition. Here we have a series of several negatives obtained by receiving the focal image of a six-inch object glass directly upon the sensitive plate, and taken, with a wide range of exposure, from five to forty seconds. Of these the one exposed the longest (fig. 85) yields a splendid and unmistakable picture of the corona, representing it, where the converging rays occurred, of a depth equal to a quarter of the Moon's diameter. Surely Dr. Gould cannot imagine the aureole of this photograph to be the chromosphere and not the corona; and yet all these pictures of Mr. Whipple's, and all of the Philadelphia expedition, and my own, agree perfectly in the features and position of the various irregularities in the outline of the corona, the difference in the representation of that object in the several photographs being solely one of extent and brilliancy. Dr. Gould adduces as an additional argument in favour of his assumption, the observation that the long coronal beams appeared to him to be "variable," while the "aureole " photographed was evidently "constant" during the time of totality. This argument,

A A

to me that although it does not suffice to answer all the questions of interest suggested by this imposing

however, loses some of its force when it is remembered that to other observers the corona appeared to the eye absolutely unchangeable, both in form and position, during the whole period of the total obscuration.' Dr. Curtis then proceeds to consider how far Dr. Gould may have been led to found his opinion upon the circumstance that the aureole' in the Philadelphia photographs, while falling far short of the height above the Moon's limb attained by the corona as seen by the eye, yet appears of very great brilliancy, rivalling the protuberances in that respect, and comes to an almost abrupt termination a short distance above the solar surface.' He shows that these peculiarities must be regarded as in all probability simply photographic effects,' the prints of the photographs 'giving every indication that the negatives from which they were taken were strongly intensified after fixing. This operation,' he adds, practised to give additional density to weak negatives, would have, in this case, precisely the effect of increasing on the photograph the apparent brilliancy of the corona without adding to its extent. Moreover, that this excessive photographic brilliancy of the under portion of the corona should not be taken as a proof of any physical or chemical peculiarity in the actual object is quite conclusively proved by my own photographs, which, while showing about the same extent of corona as those pictures of the Philadelphia party that received the least exposure, yet represent it as a very feeble laminosity, fading gradually and imperceptibly into complete darkness, and this while the same photographs show the protuberances of great brilliancy. If this peculiarity of the Burlington and Ottumwa photographs had indeed any influence in leading Dr. Gould into the misconception into which I cannot but believe he has fallen, the circumstance affords but another example among many that I have seen, of the necessity that a critic, before attempting to draw scientific inferences from photographic representations, should himself become something of a photographer, else he will be very apt to fall into this natural error of ascribing effects wholly produced in the dark room to physical characteristics of the object pourtrayed. And by a singular coincidence, evidence that Dr. Gould has not a practical acquaintance with the art would seem to be afforded in this same published letter, by his total misinterpretation of another purely photographic effect, viz., the apparent encroachment of the prominences upon the disc of the Moon as seen in the photographs. This curious appearance, instead of being due to "specular reflection," is wholly a dark-room phenomenon.'

phenomenon, it yet leaves very little room for doubt as to the general characteristics of the corona.

We are fortunately able to dispose very briefly of some of the theories respecting the corona which were suggested in old times. We need no longer inquire with close scrutiny into the theory that the corona is due to a lunar atmosphere, because we now have abundant evidence that either there is no lunar atmosphere, or that at least no atmosphere competent to produce such a remarkable appearance surrounds our satellite. We know that two very definite results (to consider no others) must inevitably follow if the Moon had an atmosphere of even moderate extent. In the first place, the refractive power of such an atmosphere would cause somewhat more than one-half of the Moon's surface to be illuminated-precisely as, in the case of our own Earth, the Sun is apparently raised by atmospheric refraction above the horizon of places lying beyond the hemisphere turned directly towards him. It is easy to show that under these circumstances, when the Moon is nearly new, her horns should extend somewhat beyond a semicircle. The fact that no such extension has been noticed suffices to prove that she has either no atmosphere or one of very limited extent. Again, the occultation of a fixed star by the Moon could not fail to be accompanied by evidence of the existence of any lunar atmosphere. Instead of disappearing suddenly, the star would be slowly reduced in brilliancy, and would appear to cling for a few moments to the outline of the Moon's disc. Since no such

appearances are noted, we must reject the conception that the Moon has an atmosphere of appreciable extent, and with it the theory, which to Kepler and Halley had seemed attractive,* that the corona is a phenomenon due to the action of a lunar atmosphere on the solar rays.

Nor need we dwell on the theory propounded by Delisle, that the corona may be an optical effect due to the diffraction of the solar rays as they pass by the Moon, because Professor Baden Powell and Sir David Brewster have abundantly demonstrated that the effects due to such diffraction could not be discernible from the Earth.

We may thus limit our attention to two general theories (each admitting of special differences) which at present divide attention. One is the theory that the corona is a solar appendage; the other is the theory that it is a phenomenon due to the passage of solar light though our own atmosphere.

It will be seen that somewhat important issues depend on the selection we have to make between these two theories. For, if the corona be but a phenomenon of our own atmosphere, it is not worthy of more attention than we might give to the rays which stream through openings between clouds and form vast beams of light across the heavens. But if it be a solar appendage, then it is one of the most imposing phenomena the

Halley mentions that contrary sentiments were entertained 'by those whose judgments he should always revere.' It has been supposed that Halley here refers to Newton.