( 359 ) 



Also of strongly represented elements some lines may appear sharp, 

 since not all lines of the same element, with given density, cause 

 anomalous dispersion in the same degree. Perhaps even there are 

 absorption lines which under no condition give rise to this pheno- 

 menon ; though this were rather improbable from the point of view 

 of the theory of light. 



Be this as it may, the mentioned limitations do not invalidate our 

 principal conclusion: that the general interpretation of the solar spectrum 

 has to be modified. We are obliged to see in Fraunhofer's lines not 

 only absorption lines, as Kirchhof does, but chiefly dispersion bands 

 (or dispersion lines). And that also on the distribution of light in 

 the stellar spectra refraction has a preponderant influence, cannot be 

 doubted either. 



We must become familiar with the idea that in the neighbourhood 

 of the celestial bodies the rays of light are in general curved, and 

 that consequently the whole interstellar space is filled with non- 

 homogeneous radiation fields ^) of different structure for the various 

 kinds of light. 



Chemistry. — "On a substance ivhich possesses numerous ^) different 

 liquid phases of lohich three at least are stable in regard to 

 the isotropous liquid." By Dr. F. M. Jaeger. (Communicated 

 by Prof. H. W. Bakhuis Roozeboom). 



§ 1. The compound which exhibits the highly remarkable phenomena 

 to be described, is cholesteryl-cinnamijlate: C^. H^j O^C .CH iCHCgHj. 



I have prepared this substance by melting together equal quantities 

 of pure cholesterol and cinnamyl-chloride in a small flask, which 

 w^as heated for about two hours in an oilbath at 190°. It is of the 

 greatest importance, not to exceed this temperature and the time of 

 heating, as otherwise the liquid mass, which commences to darken, 

 even under these conditions, yields instead of the desired derivative 

 a brown resin which in solution exhibits a green fluorescence. 



1) Das ungleichmassige Strahlungsfeld und die Dispersionsbanden. Physik. Zeitschr. 

 6, S. 239—248, 1905. 



2) In the Dutch publication, I have said : five. Since that time however, more 

 extended microscopical observation has taught me, that probably there are an 

 infinite number of anisotropous liquid phases, no sharply fixed transition being 

 observed in this manner. The hypothesis, that the transition of the first anisotropous 

 liquid phase into the isotropous should be continuous, would therefore be made 

 more probable in this way. However there are observed some irreversibilities by 

 passing from solid to li(iuid stale and vice-versa, which yet I cannot explain at this 

 moment. 



