June 15, 1905] 



NATURE 



159 



quantities of the substances. If P represents the molecular 

 weight, the product [(n— i)/d]P is the molecular refraction. 

 Landolt examined particularly the fundamental question 

 whether a different grouping of the same number of atoms 

 of the same elements — which is the cause of isomerism — 

 has any influence on the optical properties of bodies. 



He established the important fact that only the relative 

 weight of the elements is of influence on the molecular 

 refraction of a compound, while the different grouping of 

 the atoms has no appreciable effect ; and this made it 

 possible to determine the atomic refractions of the elements. 

 The atomic refraction of carbon, for instance, was obtained 

 by comparing the molecular refractions of two compounds 

 which differed only b\' one atom of carbon ; and in a 

 similar manner the atomic refractions of the remaining 

 elements were determined. 



With the aid of these constants it was now possible to 

 calculate a priori the molecular refraction of many organic 

 compounds from the elements composing them, and Landolt 

 showed that the calculated molecular refractions agreed 

 very well with those determined by experiment. 



Gladstone, in the course of his researches, was able to 

 confirm Landolt's results in many cases. But he also 

 found a considerable number of substances in which the 

 observed molecular refraction was completely at variance 

 with that obtained by adding the atomic refractions 

 together. The exceptions were so numerous that they 

 really seemed to overthrow the whole law of summation. 



Shortly before iSSo, when I was studying the literature 

 of chemical optics, a brief note published by Gladstone in 

 the Journal of the Chemical Society for May, 1870, excited 

 my attention and curiosity. The author there discusses 

 the exceptions to Landolt's rule of summation. He shows 

 firstly that in all such cases the molecular refraction is 

 never found to be too small, but always too great. Then 

 he shows that whole classes of compounds behave in this 

 abnormal fashion. 



All optically abnormal compounds proved to be rich in 

 carbon. Gladstone, therefore, examined the effect which a 

 gradual increase of carbon in the composition of a body 

 exerted on its refractivity. He found that there actually 

 was an increase in the excess of the experimental as com- 

 pared with the calculated molecular refraction, but the in- 

 crease was not regular enough to explain the anomalies. 



The saturated hydrocarbons, or paraffins, of the general 

 composition C„H.i„+2, showed normal molecular refraction. 

 Also the defines, containing two atoms less of hydrogen, 

 were found normal by Gladstone. On the other hand, the 

 hydrocarbons, containing six atoms less of hydrogen, viz. 

 the terpenes, gave molecular refractions about 3 units 

 larger than would correspond to their composition. 



With the aromatic hydrocarbons, such as benzene, 

 toluene, &c., containing eight atoms less of hydrogen, this 

 abnormal excess amounted to 6 units : — 



hydrocarbon hexane, 

 hvdrogen. 



H 



HC. 



H H \ 



HCH HCH 



CjH,,, by successive removal of 



H 



/ H \ 

 HCH HCH 



H 



-\ 



H 



(Hexan 



\c' 



H 



(Benzene) 

 CsHfi 



Paraffins 



defines 



Terpenes 



Benzene and derivatives 



.. (C„H„„+o) Normal 



,; ' -H, „ 



-He „ -f3 

 -H, „ +6 



With still further decrease in the quantity of hydrogen 

 contained (i.e. with further increase of carbon), there re- 

 sulted greater and greater refractive increments. The last 

 member of the series, however — pure carbon without any 

 hydrogen, represented by the diamond — proved to be per- 

 fectly normal in its optical properties. 



It seemed to me really extraordinarily remarkable that 

 all optically abnormal substances, without exception, gave 

 a too high molecular lefraction. It was no less astonish- 

 ing to me that the saturated hydrocarbons were optically 

 normal, but became more and more abnormal at successive 

 withdrawals of hydrogen — while pure carbon, uncombined 

 with hydrogen, is again completely normal. 



But I was most particularly struck by the quantitative 

 amount of the abnormality in the case of benzene com- 

 pounds, especially their refractive increment of six units. 

 The number 6 fascinated me. I could not help thinking 

 that therein lay the key to the mystery, and I lost no time 

 in making use of it. 



.\ccording to Kekul^'s ingenious hypothesis we can 

 imagine benzene, C^H,,, to have arisen from the saturated 



NO. 1859, VOL. 72] 



HCH HCH HCH HCH HC CH 



\H/ \H/ \^ 



\c/ \c/ ^^ 



H 



(Hexamethylene) 



Thus altogether four pairs of hydrogen atoms have been 

 removed. The elimination of the first pair was made the 

 occasion to form another simple carbon bond, like those 

 already present in he.xane, and with it the ring was closed. 

 The splitting-off of the other three pairs of hydrogen atoms, 

 on the other hand, resulted in the formation of three double 

 bonds of carbon atoms — a kind of bond which does not 

 occur in the optically normal hexane. 



Now Gladstone had found that benzene exhibits a re- 

 fractive increment of 6 units. Reading this, I was struck 

 in a moment by the thought : might not this abnormal re- 

 fractive increment of benzene be due to its double carbon 

 bonds, which are absent in the optically normal hexane? 

 If this were so, I went on to reason, since three double 

 bonds in benzene correspond to a refractive increment of 

 6 units, therefore one double bond must entail the incre- 

 ment of 2. 



These ideas received no support whatever from the then 

 known facts. For Gladstone had stated expressly that the 

 olefines, i.e. open-chain hydrocarbons, containing one double 

 carbon bond, were optically normal. However, I did not 

 allow myself to be discouraged ; and my expectations were 

 confirmed by the very first experiment. The olefine ex- 

 amined not only proved to be optically abnormal, but gave 

 the predicted refractive increment of 2 units, corresponding 

 to the presence of one double carbon bond. Gladstone, 

 therefore, as I had supposed, was mistaken in this case. 

 Further experiments proved that not one of the olefines 

 was optically normal. Without exception they gave the re- 

 fractive increment of 2 units, one-third of that of benzene. 



I next proceeded to examine the di-olefines — substances 

 which contain tivo double carbon bonds. Here also, in 

 conformity with expectation, a constant refractive incre- 

 ment was found, double as large as that of the olefines and 

 two-thirds of that of benzene : — 



Paraffins 



Olefines 



Diolefines 



Benzene compounds. 



(C„H.,„+J 



Normal 

 -H., „ +2 

 -H4 ,, +4 

 H, ,. +6 



The dimensions of our subject this evening prevent the 

 detailed demonstration of these important facts by experi- 

 ment. I will only show you that the spectrum of a 

 saturated hydrocarbon (a paraffin) is distinguishable at a 

 glance from that of a substance containing double bonds. 



On this screen we project the electric spectrum of metallic 

 calcium. First we cause the rays of light to pass through 

 a prism filled with paraffin oil. Then we e.xchange this 

 prism for another, filled with a substance containing atoms 

 linked by double bonds. (Experiment.) 



In the second case you observe, first, a much greater 

 deviation of the whole spectrum, i.e. greater refraction, 

 and secondly, far wider intervals between the coloured 

 lines of the spectrum, i.e. greater dispersion, which is 

 usually correlative to the refraction. 



Thus quantitative experimental confirmation was obtained 

 for the view that abnormal refractive increments which 

 increase with the diminution of hydrogen contained in the 

 substances are caused by the presence of double carbon 

 bonds. 



.At the same time, however, the experiments yielded a 

 second result of fundamental importance. The olefines 

 contain 2, and the diolefines 4, atoms of hydrogen less than 

 the paraffins. Similarly the refractive increment of the 

 olefines is 2, and of the diolefines 4. 



Benzene, C^H,, contains 8 atoms of hydrogen less than 

 the corresponding paraffin, hexane, C^H,,. The increment 

 of benzene, however, amounts not to 8, but to 6 ! Thus 



