586 



NATURE 



[Oct. 12, 1882 



In the same year I made a series of experiments on the other 

 magnetic metal, nickel and cobalt, and found that whilst cobalt 

 lengthened under magnetisation, nickel appeared to suffer no 

 change. 1 This result is surprising, for nickel more nearly 

 resembles iron and cobalt thin steel in magnetic properties, the 

 formei having little coercive force, and the latter very consider- 

 able retentive power. With entirely new apparatus the experi- 

 ments were repeated, and a distinct shortening of the nickel was 

 now found, cobalt elongating but not so much as iron. This 

 observation is, I believe, new, the fact was first noticed by me 

 on September 9, 1873, t> ut some uncertainty as to the reliability 

 of the apparatus I then used led me to put the matter aside till 

 July, 1876, when the experiments were repeated, and the fact 

 that cobalt elongates and nickel retracts under magnetisation, 

 was fully confirmed. 



The multiplying apparatus that was found to yield most satisfac- 

 tory results was a simple form of optical lever, a mirror vertically 

 fixed over the fulcrum of a lever of the first order, and reflecting 

 a scale at some distance into an observing telescope. The appa- 

 ratus will be more fully described in the report that will be pre- 

 sented next year ; a committee, with a small money grant, 

 having been appointed at a previous meeting of the Association 

 to investigate this and certain other molecular changes accom- 

 panying the magnetisation of iron, described by the author at th.2 

 Bradford meeting of the Association. 



The results so far obtained may be summed up as follows : — 

 However often the current traverses the helix around the bar of 

 cobalt, the elongation is practically the same after the first 

 current, and amounts to about two-thirds of the elongation pro- 

 duced in an iron bar of the same dimensions. In my measure- 

 ment the elongation of the iron amounted to about l-26o,oooth 

 of its length for the maximum magnetisation ; the iron elongated 

 5 scale divisions, and the cobalt 3, or 1-425, oooth of its length. 

 With nickel, the retraction on the same scale was 10, or twice 

 the elongation of the iron, or about 1-130, oooth of the length of 

 the bar. Reversing the current made no alteration in the results. 

 The index returned promptly to zero on the cessation of the 

 current. The retraction of the nickel was so instantaneous that 

 it was only by noting the scale-reading that any motion could 

 be discovered to have taken place. The helix in all ca-es was 

 the whole length of the bars. 



Inclosing the bars in a vessel of water terminating in a capil- 

 lary tube (the stem of a mercurial thermometer of extremely fine 

 bore), and surrounding the vessel by a powerful magnetising 

 helix, no motion of the water-level in the capillary tube was 

 noticed with iron and cobalt on the making, breaking, or re- 

 versing the current in the helix ; with nickel no motion was 

 observed on making, and a barely perceptible, but still definite, 

 fall of the index, equal to about l-io,ooo,oooth of the volume 

 of the bar, occurred on breaking, w hich was more cleirly seen 

 by frequent interruptions of the current. 



The " magnetic tick " is heard loudly with cobalt and nickel, 

 as well as iron, the former giving a very clear metallic click on 

 magnetisation. 



I am much indebted to the kindness of Messrs. Johnson and 

 Matthey for the bars of nickel and cobalt (oi inches long and 1 

 inch diameter) with which the experiments were conducted, and 

 also to Mr. Gore, F.R.S., for the loan of a longer bar of nickel. 

 Experiments are now in progress to determine the effect of tem- 

 peratures and longitudinal tensions on the result. 



Preliminary experiments show, that raising the temperature 

 of the iron and cobalt bars some 50° C. makes a scarcely appre- 

 ciable difference in the amount they elongate, whereas, when 

 nickel is heated the same amount, its retractio n on magnetisa- 

 tion is, as might be expected, considerably diminished, being 

 about three-fourth; of the amount occurring at the temperature 

 of the air. Owing to the short length of the bars, the actual 

 elongation measured was, in the case of the cobalt, only the 

 I-46,oooth of an inch, but a difference of 100,000th of an inch 

 could confidently be measured. 



SUNLIGHT AND SKYLIGHT AT HIGH ALTI- 

 TUDES 



AT the Southampton meeting of the British Association, 

 Captain Abney read a paper in which he called atten- 

 tion to the fact that photographs taken at high altitudes show 

 s kies that are nearly black by comparison with bright objects 

 1 Phil. Mag., January, 1874. 



projected against them, and he went on to show that the 

 higher above the sea-level the observer went, the darker the 

 sky really is and the fainter the spectrum. In fact, the latter 

 shows but little more than a band in the violet and ultra- 

 violet at a height of 8500 feet, whilst at sea-level it shows 

 nearly the whole photographic spectrum. The only reason of 

 this must be particles of some reflecting matter from which sun- 

 light is reflected. The author refers this to watery stuff of which 

 nine-tenths is left behind at the altitude at which he worked, 

 lie then showed that the brightness of the ultra-violet of direct 

 sunlight increased enormously the higher the observer went, but 

 only 10 a certain point, for the spectrum suddenly terminated 

 about 2940 wave-length. This abrupt absorption was due to 

 extra atmospheric causes and perhaps to space. The increase in 

 brightness of the ultra-violet was such that the usually invisible 

 rays L, M, N could be distinctly seen showing that the visibility 

 of these rays depended on the intensity of the radiation. The 

 red and ultra-red part of the spectrum was also considered He 

 showed that the absorption lines were present in undiminished 

 force and number at this high altitude, thus placing their ori'in 

 to extra atmospheric causes. The absorption from atmospheric 

 causes of radiant energy in these parts he showed was due to 

 'water-stuff," which he hesitated to call aqueous vapour, since 

 the banded spectrum of water was present, and not lines. The 

 B and A line he aLo stated could not be claimed as telluric lines, 

 much less as due to aqueous vapour, but must originate between 

 the sun and our atmosphere. The author finally confirmed the 

 presence of benzene and ethyl in the same region. He had found 

 their presence indicated in the spectrum at sea-level, and found 

 their absorption lines with undiminished intensity at 8500 feet. 

 Thus without much doubt hydrocarbons must exist between our 

 atmosphere and the sun, and it may be in space. 



Prof. Langley, following Capt. Abney, observed : The 

 very remarkable paper just read by Captain Abney has already 

 brought information, upon some points which the one I am 

 about, by the courtesy of the Association, to present, leaves in 

 doubt. It will be understood then that the references here are 

 to his published memoirs only, and not to what we have 

 just heard. 



The solar spectrum is so commonly supposed to have been 

 mapped with completeness, that the statement that much more 

 than one half its extent is not only unmapped but nearly un- 

 known, may excite surprise. This statement is, however, I 

 think, quite within the truth, as to that almost unexplored region 

 discovered by the elder Herschel, which lying below the red 

 and invisible to the eye, is so compressed by the prism, that 

 though its aggregate heat effects have been studied through the 

 thermopile, it is only by the recent researches of Capt. Abney 

 that we have any certain knowledge of the lines of absorption 

 there, even in part. Though the last named investigator has 

 extended our knowledge of it to a point much beyond the lowest 

 visible ray, there yet remains a still remoter region, more exten- 

 sive than the whole visible spectrum, the study of which has 

 been entered on at Alleghany, by means of the linear Bolometer. 

 The whole spectrum, visible and invisible, is powerfully 

 affected by the selective absorption of our atmosphere, and that 

 of the sun ; and we must first observe that could we get outside 

 our earth's atmospheric shell, we should see a second and very 

 different spectrum, and could we afterward remove the solar 

 atmosphere also, we should have yet a third, different from 

 either. The charts exhibited show : — 



1st. The distribution of the solar energy as we receive it, at 

 the earth's surface, throughout the entire invisible as well as 

 visible portion, both on the prismatic and normal scales. This 

 is what I have principally to speak of now, but this whole first 

 research is but incidental to others upon the spectra before any 

 absorption, which though incomplete, I wish to briefly allude to 

 later. The other curves then indicate. 



2nd. The distribution of energy before absorption by our 

 own atmosphere. 



3rd. This distribution at the photosophere of the sun. 

 The extent of the field, newly studied, is shown by this draw- 

 ing (chart exhibited). Between H in the extreme violet, and A 

 in the furthest red, lies the visible spectrum, with which we are 

 familiar, its length being about 4,000 of Angstrom's units. If, 

 then, 4,000 represent the length of the visible spectrum, the 

 chart shows that the region below extends through 24,000 more, 

 and so much of this as lies below wavelength, 12,000, I think, 

 is now mapped for the first time. 



Wehaveto A.= 12,000, relatively complete photographs, pub- 



