i6 



NATURE 



[November i, 1894 



nearly two degrees, or 7200 seconds, and that we have 

 just seen by the illustration of the D lines that lines 

 3 inches apart can be thus divided, we may see for 

 ourselves that, at any rate, over 2000 lines, if they 

 exist, can be mapped. But these lines do exist, the 

 whole of this new region being apparently as inti- 

 mately tilled by them as the visible spectrum by the 

 Fraunhofer lines. In further evidence of this, here 

 is a portion of the lower spectrum in the comparatively 

 unknown part extending from X = r4/i to \ = i'2\i. in- 

 cluding the great band 2 shown as a single inflection in 

 my first communication to this Association, but here re- 

 solved into thirty or more subordinate lines (Fig. 3). This 

 illustration includes a part of the new region discovered 

 on Mount Whitney in iSSi ; and in the small portion 

 here exhibited, you may see that about 200 lines are 

 discriminated. 



I am now trj-ing to bring what may be called the 

 first stage of the long labour, a portion of which is 

 here described, to a close, this first stage consisting 



the expense of the invisible, nor even on such a log- 

 arithmic one as that proposed by Lord Rayleigh, but on a 

 conventional scale, which 1 will ask you to tolerate, as it 

 is simply meant to show the actual extent and importance 

 of the region covered here as compared with that known 

 to Newton. In this illustration, with which I close my 

 remarks, the mean dispersion throughout the invisible 

 rock-salt spectrum, as far as 4^1, is taken as the standard, 

 and both spectra are laid out on that common scale. On 

 the left is the visible spectrum known to Newton ; next 

 this, is the region known through photography, now ex- 

 tending a little beyond the band, /jot, which marks what 

 at the time these researches were commenced, was con- 

 sidered by the then most distinguished investigator, in 

 the infia-red, the end of the heat spectrum. Beyond, 

 and on the right, is a part of the new regions of the spec- 

 trum developed by the bolometer, and of which charts 

 may be shortly expected on the scale of which a speci- 

 men in detail has just been shown. 



1 cannot close this statement without expressing the 



I 



I I ! 



lii 



KiG. 3. — iiulograph of the portion of the infra-red Eolar spectrum lying telw.-cn wave-lengths i j/x and 2*?fA. 



chiefly in the discovery of, and mapping the relative 

 positions of new spectral lines. 



I will only refer to what it seems to me the second 

 part of this work is likely to be, and to the different kind 

 of interest which may not improbably belong to it, from 

 that which belongs to this, the first. 



We are thus far in the position of early students of the 

 visible spectrum, who simply drew the lines they saw, 

 without inquiring into their meaning. Nevertheless, to 

 have discovered and mapped a great number of these 

 lines is only a beginning, for their real value lies in 

 their interpretation, and this is still chiefly to come, | 

 As to the possible importance of this interpreta- 

 tion, it is not enough to remind ourselves that three- 

 quarters of the whole energy of the sun exists here, 

 and not in the upper spectrum. We must remember 

 also that while, as a rule, in the upper and visible spec- 

 trum a great proportion of the lines are caused by 

 absorption in the solar atmosphere, and a perhaps smaller 

 portion by telluric absorption, here, on the contrary, we 

 are led. by everything we already know, to expect that 

 the great telluric absorptions on which meteorological 

 predictions and other immediately practical interests 

 depend, may be expected to be found, and it is on the 

 comparison of these energy curves taken at different 

 periods of the year, and at different altitudes of the sun, 

 that those who are engaged in the work see good cause 

 to hope for important results in the future. 



Before I conclude, let me present a collective view of 

 the field in which work has been going on in these 

 later years at the Smithsonian Observatory, on the same 

 scale, with the visible spectrum. I say "on the same 

 scale," meaning, not on a wave-length scale, which ex- 

 pands the invisible at the expense of the visible, and not 

 on a prismatic scale alone, which expands the visible at 



NO. 1305, VOL. 51] 



gratification with which I have laid it before the s.-ime 

 body that listened to that made on the same subject 

 twelve years ago, or my sense of my good fortune, in 

 doing so before an audience in which 1 recognise many 

 of the same eminent men who so kindly received that 

 first presentation of these researches. 



THE TREA TMENT OF DIPHTHERIA 

 ANTI-TOXIC SERUM. 



BY 



FOUR years ago Prof. Behring published his remark- 

 able paper '' On the mechanism of immunity ag:ainst 

 experimental diphtheria in animals.'' In this memoir the 

 author stated that it was possible to immunise animals 

 against the diphtheria b.icillus by the injection of cultures 

 attenuated by heat or the addition of i in 500 tri- 

 chloride of iodine to the cultivating medium. The same 

 result could be obtained by the inoculation of the pleural 

 exudation of animals dead of experimental diphtheria, 

 or by the injection of chemical compounds, such as 

 trichloride of iodine, after inoculation of virulent diph- 

 theria-bacilli. 



liehring's most important discovery, however, was 

 that the serum of animals immune against the bacillus 

 of diphtheria and its poisons had the power of " destroy- 

 ing" in vitro and in the animal body the chemical poisoix 

 secreted by this bacillus ; and that animals, after a mortal 

 dose of <ii|)htheria poison had been injected, could be not 

 only immunised, but actually cured, by the introduction 

 into their system of the serum of animals immunised 

 against the specific bacillus and its poisons. In a further 

 series of researches he found that the serum of such 

 animals possessed this power to a most remarkable and 



