with the Chemical Constituents of the Solar System. 381 



having a greater molecular weight. This is accompanied by 

 other metallic Yapours of a similar degree of lightness, which 

 probahlv condense to form that region of cloud which consti- 

 tutes the glowing surface that we behold. The spectroscope 

 has shown ihe absence in the solar atmosphere of mercury, 

 tellurium, bismuth, and antimony, bodies of high molecular 

 weight, which have been detected in Aldebaran — and, indeed, 

 of all the metals of high atomic weight, except perhaps barium 

 and traces of lead. At the same time we find no indication of 

 any of the yery infusible metals, such as platinum or gold. 

 As to the centre of the sun itself, we are of course perfectly 

 ignorant of its chemical constitution. 



II. Let us now suppose the whole solar system to have been 

 a great revolving nebula condensing to the central sun, and 

 forming from its outer portion such globes as the planets and 

 their satellites, or such accumulations of smaller particles as 

 the comets, meteorites, the group of asteroids, or the rings of 

 Saturn. How may we expect to find these bodies constituted ? 

 Clearly we may expect to find that they consist principally of 

 those elementary substances which give the lightest or the least 

 condensable vapours, while at the same time we may expect 

 them to contain smaller quantities of the less volatile and of 

 the heavier ones. !Now the earth, though it is one of the pla- 

 nets nearest the sun, must still be regarded as formed from 

 the external part of the great nebula, when we bear in mind 

 the immense mass of the central sun itself. It became of in- 

 terest to see whether the relative quantities of the known 

 chemical elements in the earth was in conformity with this 

 hypothesis. I asked two of my assistants to draw out a list of 

 the elements, dividing them according to their abundance or 

 otherwise on the surface of the globe. I did not construct 

 this Table myself, lest in any instances my judgment should 

 be warped by the theory which it was intended lo test. The 

 following is the Table which they drew up: — 



Elements and Vapour-densit'es. 



N"on-metallic Elements. ' 



Not plentiful. 



Plentiful. 



Oxygen 16 



Silicon 28-5 



Carbon 12 



Hvdrogen 1 



Siilphur 32 



Chlorine . . 35-5 



Nitrog-en 14 



Averao-e .... 19-8 



Phosphorus. ....... 62 



Fluorine 19 



Bromine 80 



Boron 11 



Iodine 127 



Selenium 79 



Average 63 



