metal, depends upon such a small matter as one line among 10,000, what will happen if á man neglects to observe this change? People will say, "Oh! in a research of that kind it is altogether excusable if he has made a mistake." But if we have a series of phenomena recorded by means of a camera on "a retina which never forgets," as Mr. Delarue has beautifully put it, and if we compare those pictures day by day and year by year, the thing is put beyond all question when we observe one line disappearing, or another line appearing.

I have already given a drawing of the part of the solar spectrum near the line H. I wish to call attention to one line in this region. We have an admirable map of the solar spectrum made about the year 1860. The draughtsman, recording by means of his eye the lines in the spectrum, would not be very likely to overlook a line darker than some he inserts, but he might easily overlook finer lines. Now, it is a fact that in the most careful map that we have a map drawn with a most wonderful honesty and splendid skill—a line is absent in the region indicated, which line is now darker than some that were then drawn, and that line indicates the presence of an additional element in the sun-strontium.

I do not make this assertion thinking that subsequent facts will show the drawing to be wrong, but because I see reason to believe that what we know already of the sun teaches us that it is one of the most likely things in the world that strontium was not present in such great

quantity in the reversing layer when the drawing was made; but, however that may be, all must recognize how important it is that photographs should be compared with photographs made five, ten, fifteen, a hundred, or two hundred, or as many years as you like apart, and it is in this possible continuity of observation of the solar spectrum, carried on for centuries, that I think we have in photography not a tremendous ally of the spectroscope, but a part of the spectroscope itself. Spectroscopy, I think, has already arrived at such a point, at all events in connection with the heavenly bodies, that it is almost useless unless the record is a photographic one.

II. Stellar Photography.

Dr. Draper and Dr. Huggins have succeeded in getting photographs of the spectra of some of the stars. This is a matter of the very highest importance, because the sun is nothing but a star, and the stars are nothing but distant suns; and as long as we merely investigate our sun, however diligently or admirably we do it, and neglect all the others, it is as if a man who might have the whole realm of literature to work at should confine himself to one book, and that book probably not a good representative of the literature of the country he was inquiring into.

To the spectroscope all nature is one, and it is absolutely impossible to make a single observation, either on a sun, or a star, or a comet, without bringing chemical and

physical considerations into play; and it will be a regrettable circumstance if chemists employ the spectroscope in terrestrial chemistry—they have not done much in that way yet-without taking the sun and all the various stars of heaven into counsel, because the spectroscope is absolutely regardless of space, and tells us that the elements which are most familiar to us here, or at all events a good many of them, are present in the most distant stars, and the spectroscope shows us those elements existing under conditions which further are absolutely impossible here.

§ 12. The Work of the Future.

We have, I believe, what we may almost call a new chemistry, some day to be revealed to us by means of photographic records of the behaviour of molecules. Recollect that the difference between the iron spectrum of one line and the iron spectrum of between 400 and 500 lines is simply due to the difference in the vibration of the molecules or atoms of iron in the centre of the electric arc and its exterior. There is one question which all students of the spectroscope may ask of photographers, and it is this. Why should we any longer be confined, in registering spectra, to the more refrangible end of the spectrum, when one of the very first spectra of the sun that was ever taken was a complete photograph of the spectrum, including not only the blue and the green, but the yellow, the red, and the extreme red? I think that if photographers will study

the action of light on molecules, and will give those who are familiar with the spectroscope, and those who are anxious to promote the progress of spectroscopic research, a means of extending photographic registration, not only into the green part of the spectrum which they do already with difficulty, but to the extreme red, then the use of the eye will almost entirely be abolished in these inquiries. And although no one has a higher estimate than myself of the extreme importance of the eye, I think that the more it is replaced by permanent natural records in these inquiries, the better it will be for the progress of Science.

As I have already hinted, I believe the spectroscope is capable of leading along the track of discovery in this matter. It has begun, I think, by altogether abolishing the distinction drawn in all our text-books between the chemical and the visible rays. The curves which are given in these books, showing us the maxima of heat, light, and chemical action, are, I fancy, merely curves showing us, as it were, the absorption spectra of those substances by which the maxima have been determined—whether they be lamp-black, the coating of the retina, or salts of silver, and are really altogether independent of the nature of light. The salts of silver ordinarily employed require short waves to set them vibrating first, and dissociating afterwards as a result of the vibration. Why should not other salts shiver even under the influence of the ultra red? What it is that renders a molecule apt to vibrate with one wave-length more than another remains to be discovered; there is