September 12, 1884.] 



SCIENCE 



255 



particles lit up in succession at this rate. This would 

 be equivalent to a shortening of the wave-length of 

 light by about tWotttt part; and, in the case of the 

 yellow sodium lines, would produce a shift of a dis- 

 tance of about tot of the space between the two 

 lines. It would require an instrument of very high 

 diffusive power and sharply denned lines to make 

 such a displacement appreciable. With lines of 

 longer wave-length, the displacement would be pro- 

 portionately greater; while, if a receding explosion 

 could be observed simultaneously with an advancing 

 one, the relative shift would be doubled. In this 

 way the two images of the red lithium line would be 

 separated by about ^ of a unit of Angstrom's scale, 

 a distance about equal to that between the compo- 

 nents of the less refrangible of the pair of E lines. 



The experiments were made first in a straight glass 

 tube, and then in a U tube, which enabled them to 

 observe the advancing and retreating wave. In these 

 cases it was found that the calcium spectrum was pro- 

 duced, owing to particles of the glass detached by the 

 explosive reaction. The reversals showed too, that, in 

 the wave of explosion, the gases do not reach their 

 maximum temperatures all at once, but the front of 

 the wave is cooler than the part which follows and 

 absorbs some of its radiations, while the rear of the 

 wave does not produce the same effect. 



Experiments were now made in iron tubes, and 

 here the spectrum of iron was obtained from the 

 particles detached from the tube. Altogether, sixty- 

 eight lines of iron were identified, of which about 

 forty lie in the ultra-violet between hydrogen and 

 oxygen. Only one iron line above oxygen was defi- 

 nitely seen, and that in only a few photographs. 

 Since iron gave so many lines, linings of copper, 

 lead, cadmium, zinc, aluminium, and tin were in- 

 serted in the tube. Cadmium, aluminium, and tin 

 gave no lines whatever; zinc gave only a doubtful 

 impression; lead gave one visible line, and two in 

 the ultra-violet; copper gave one visible line in the 

 green, two in the ultra-violet, and occasionally a 

 shaded band in the blue; cobalt and nickel gave a 

 great many lines. Berthelot and Yielle having put 

 the temperature produced by the explosion of hydro- 

 gen and oxygen under a pressure of 9.8 atmospheres 

 at 3,240°, the authors believe that they cannot be far 

 wrong in assuming the temperature at about 3,000°, 

 and that at this temperature such metals as iron, 

 nickel, and cobalt are vaporous, and emit many char- 

 acteristic rays, and that by far the greatest part of 

 these rays lie between G and P. 



The discussion on the constitution of the elements 

 was opened by Dr. Dewar ; and after referring to the 

 doctrine of continuity found in the essays of Grove, 

 taught by Black, and held by Xewton, and the views 

 of Clerk Maxwell who said that the process by which 

 atoms are formed cannot be known, since they are 

 neither born nor do they die, he stated that our 

 recent knowledge on the constitution of molecules 

 was largely due to the studies of Deville upon disso- 

 ciation, and that he was led to make these studies 

 from the observation of Grove that a platinum bead 

 heated in an oxyhydrogen flame would decompose 



water when immersed in it. Experiments made by 

 Dr. Dewar in this direction were described ; and it was 

 stated that chemical bodies are not fixed or unstable 

 at certain fixed temperatures, but that there exists a 

 relation between the pressure, temperature, and char- 

 acter of the body, which determined its stability. De- 

 ville held the change to be similar to a change in state 

 of bodies ; and, this relation being true, thermodynam- 

 ics enable us to determine the amount of change 

 for given conditions. The change of state in elemen- 

 tary substances is not unlike a chemical change. 



The spectroscope has been used to study the con- 

 stitution of molecules, and Roscoe has found that 

 the allotropic forms of bodies give different spectra. 

 Lockyer has attempted to show the evolution of the 

 elementary bodies from hydrogen. His results have 

 been criticised as having been due to the presence 

 of impurities, but Lockyer disproved this. It has 

 been said, too, that he did not use a spectroscope of 

 sufficiently dispersive power. 



Prout's hypothesis was next considered; and it was 

 shown that the most careful determinations of the 

 atomic weights of nitrogen, potassium, magnesium, 

 zinc, and bismuth, by Stas and Marignac, yielded re- 

 sults that were not simple multiples of hydrogen. 



In continuing the discussion, Dr. Gibbs said he was 

 not sure that the accepted views of the molecular 

 constitution of chemical compounds was the correct 

 one. Taking common salt, for instance, it might in 

 the solid state be composed of one hundred molecules 

 of sodium and one hundred molecules of chlorine; 

 when in solution it might be simpler; when in the 

 gaseous form, simpler still; and when exposed to a 

 vacuum, such as Mr. Crookes has produced, it might 

 have the accepted constitution. He referred to the fact, 

 that Professors Liveing and Dewar had found that 

 cadmium, mercury, and zinc gave no spectra at high 

 temperatures. As these are all monatomic molecules, 

 it might be that in this process we possessed a means 

 for studying the constitution of the elementary mole- 

 cules. Professor Liveing thought Dr. Gibbs' s sug- 

 gestion concerning the action of the monatomic 

 elements an improbable one, since aluminium and 

 tin gave no lines under the same conditions. He 

 said that many lines of iron suggested either a very 

 complex constitution, or else that the substance we 

 term iron is really formed of a number of elements 

 which yet defy separation, and which have nearly 

 similar atomic weights. We have an instance of 

 such a case in the cerium group. The D 3 line, for 

 instance, may belong to an element more volatile 

 than hydrogen. Sir Lyon Playfair pointed out, that 

 when solid iodine was immersed in liquid sulphur- 

 ous acid no action resulted ; but if the iodine was in 

 solution, and the sulphurous acid gaseous, they com- 

 bined readily. He suggested the study of the tem- 

 perature at which iodine or sulphur would combine 

 with sulphurous acid. Dr. Tilden said we needed 

 more extensive and accurate observations on the 

 temperature at which chemical action — such, for 

 instance, as the point of ignition — begins. Dr. 

 Dewar stated that we have, in the result of the 

 researches of Dr. Perkins in the magnetic rotation of 



