equal and opposite (+); the tangential plane at that point is y-40+3=0.

1

8

The border-curve may be considered as a twisted curve lying on this surface; it starts from the point (, 0, 0), where it touches the line y=0, -=0; it touches also the line x 3 81 y=0=1 at the point (1, 1, 1), and touches the axis of a at infinity.

We discussed in the previous paper the equation of this curve, and the practical method of obtaining points on it, such a series of points is given in Table IV. (owing to the method of obtaining them, the values are only approximate).

lines perpen1 and 3 of the The projection

It is the projections of this curve on the dicular to the axes which are shown in figs. present paper and fig. 2 of my previous paper. in fig. 1 touches the axis of where 0=0, and the line y-40+3=0 at the point (1, 1); the projection in fig. 3 touches that curve of the family there shown for which y=0 at the point, 0) (and makes an angle tan-1 with the axis of a there); it also touches the line =1 at (1, 1), and is asymptotic to the axis of x.

THE

VI. The New Spectrum. By S. P. LANGLEY *.

[Plate III.]

HE writer (at the concluding meeting of the National Academy of Sciences on April 18) remarked on the disadvantages in the matter of interest of the work of the physicist, which he was about to show them, to that of the biologist, which was concerned with the ever absorbing problem of life. He had, however, something which seemed to him of interest, even in this respect, to speak of, for it included some indications he believed to be new, pointing the way to future knowledge of the connexion of terrestrial life with that physical creator of all life, the sun.

He had to present to the Academy a book embodying the labour of twenty years, though at this late hour he could. scarcely more than show the volume with a mention of the leading captions of its subject. What he had to say then would be understood as only a sort of introductory description of the contents of the work in question, which was entitled "Volume I. of the Annals of the Astrophysical Observatory of the Smithsonian Institution."

In illustration of a principal feature of this book, the Academy saw before them on the wall an extended solar spectrum, only a small portion of the beginning of which, on the left, was the visible spectrum known to Sir Isaac Newton. This was the familiar visible coloured spectrum which we all have seen and know something of, even if our special studies are in other fields.

It is chiefly this visible part which has been hitherto the seat of prolonged spectroscopic investigation, from a little beyond the violet, at a wave-length of somewhat less than 04", down to the extreme red, which is generally considered to terminate at the almost invisible line A, whose wave-length is 0-76". On the scale of the actual wave-length of light, then, where the unit of measurement (1") is one one-thousandth of a millimetre, the length of the visible spectrum is 0.36".

The undue importance which this visible region has assumed, not only in the eyes of the public, but in the work of the spectroscopist, is easily intelligible, being due primarily to the evident fact that we all possess, as a gift from nature, a

* Abstract of a paper read before the National Academy of Sciences at its Washington meeting, April 18, 1901. From an advance proof communicated by the Author. We are indebted to the author for the plate accompanying this paper.

wonderful instrument for noting the sun's energy in this part, and in this part only.

While, then, this part alone can be seen by all, yet the idea of its undue importance is also owing to the circumstance that the operation of the ordinary prism gives an immensely extended linear depiction of the really small amount of energy in this visible part. There is also a region beyond the violet, most insignificant in energy and invisible to the eye, and the association of this linear extension due to the prism, with the accident that the salts of silver used in photography are extraordinarily sensitive to these short wave-length rays, so that they can depict them even through the most extreme enfeeblement of the energy involved in producing them, also makes this part have undue prominence. This action of the prism and of the photograph is local, then, and peculiar to the short wave-lengths; and owing to it, all but special students of the subject are, as a rule, under a wholly erroneous impression of the relative importance of what is visible and what is not. The spectrum has really no positive dimension, being extended at one end or the other according to the use of the prism or grating employed in producing it. Perhaps the only fair measurement for displaying a linear representation of the energy would be that of a special scheme, which the writer had proposed, in which the energy is everywhere the same; but this presentation is unusual and would not be generally intelligible without explanation.

The

The map before us (Pl. III.) will be intelligible when it is stated that it is, as to the infra-red, an exact representation of that part of the spectrum given by a rock-salt prism. visible and ultra-violet spectrum given here is not exact, for the reason that it would take nearly a hundred feet of map to depict it on the prismatic scale, though this is caused by but a small fraction of the sun's energy; so monstrous is the exaggeration due to the dispersion of the prism.

Looking, then, at the map: First, in the spectrum on the left and beyond 0'4" is the ultra-violet region, in fact almost invisibly small, but which in most photographs shows almost a hundred times larger than the whole infra-red. It really contains much less than one hundredth part of the total solar energy which exists. Beyond it is the visible spectrum, containing perhaps one fifth of the solar energy.

As the writer has elsewhere said, "the amount of energy in any region of the spectrum, such as that in any colour, or between any two specified limits, is a definite quantity, fixed

* American Journal of Science [3] xxvii. p. 169 (1884).

by facts, which are independent of our choice, such as the nature of the radiant body or the absorption which the ray has undergone. Beyond this Nature has no law which must govern us."

Everything in the linear presentation, then, depends on the scale adopted. In other words, if we have the lengths proportionable to the energies, the familiar prismatic representation enormously exaggerates the importance of the visible, and still more of the ultra-violet region, and similarly the grating spectrum exaggerates that of the infra-red region. Now he had given, on the map before them, and through the whole infra-red, the exact rock-salt prismatic spectrum, but for the purpose of obtaining a length which represented (though insufficiently) that of the visible spectrum, he had laid the latter down on the average dispersion in the infrared, which was perhaps as fair a plan as could be taken for showing the approximate relation of the two fields of energy in an intelligible way, though it gave the visible energy too small.

Lt us recall, then, at the risk of iteration, that in spite of the familiar extended photographic spectra of the hundreds of lines shown in the ultra-violet, and in those of the coloured spectrum, it is not here that the real creative energy of the sun is to be studied, but elsewhere, on the right of the drawing, in the infra-red. Looking to the spectrum as thus delineated, next to the invisibly small and weak ultra-violet, comes the visible or Newtonian spectrum, which is here somewhat insufficiently shown, and on the right extends the great invisible spectrum in which four-fifths of the solar energies are now known to exist.

Of this immense invisible region nothing was known until the year 1800*, when Sir William Herschel found heat there with the thermometer.

After that little was done + (except an ingenious experi

Philosophical Transactions, vol. xc. p. 284 (1800).

+ It should, however, be mentioned that an important paper by Draper (Lond. Ed. & Dublin Phil Mag., May 1843) was published in 1843, in which he appears to claim the discovery of the group here called por and which is now known to have a wave-length of less than 1". (Its true wave-length was not determined till much later.) Later, Fizeau seems to have found further irre.ularities of this heat as long ago as 1847, and of its location, obtaining his wave-lengths by means of interference-bands. His instrumental processes, though correct in theory, were not exact in practice and yet it seems pretty clear that he obtained some sort of lecognition of a something indicating heat, as far down as the great region immediately above 2 on our present charts. Mouton (Comptes Rendus, 1879) confirmed this observation of Fizeau's and contrived to get at least an approximate wave-length of the point where the spectrum (to him) ended, at about 1.8".

ment by Sir John Herschel * to show that the heat was not continuous) till the first drawing of the energy curve by Lamansky +, in 1871, which, on account of its great importance in the history of the subject, is given on the map. It consists of the energy curves of the visible spectrum, and beyond it, on the right (and in illustration of what has just been said it will be seen how relatively small these latter appear), of three depressions indicating lapses of heat in the infra-red. It is almost impossible to tell what these lapses are meant for, without a scale of some kind (which he does not furnish), but they probably indicate something, going down to near a wave-length of 1". It is obvious that the detail is of the very crudest, and yet this drawing of Lamansky's was remarkable as the first drawing of the energy spectrum. It attracted general attention, and was the immediate cause of the writer's taking up his researches in this direction.

It seems proper to state here that the true wave-lengths were at that time most imperfectly known, but that in 1884, and later in 1885, they were completely determined by the writer as far as the end of what he has called "the new spectrum" at a wave-length of 5.3".

The upper portion of the infra-red is quite accessible to photography, and the next important publication in this direction was that of Captain (now Sir William) Abney §. which gave the photographic spectrum down to about 1·1", much beyond which photography has never mapped since.

From the time of seeing Lamansky's drawing, the writer had grown interested in this work, but found the thermopile, the instrument of his predecessors, and the most delicate then known to science, insufficient in the feeble heat of the grating spectrum, and about 1880 he had invented the bolometer || and was using it in that year for these researches. This may perhaps seem the place to speak of this instrument, though with the later developments which have made it what it is to-day, it has grown to something very different from what it was then.

It has, in fact, since found very general acceptance among physicists, especially since it has lately reached a degree of

* Philosophical Transactions, vol. cxxx. p. 1, 1840.

+ Monatsberichte der k. Akademie der Wissenschaften zu Berlin, December 1871.

American Journal of Science, March 1884, and August 1886.

§ Philosophical Transactions, vol. clxxi. p. 653, 1880.

||| Actinic balance, American Journal of Science, 3rd series, vol. xxi. p. 187, 1881.