On the Constituents of Cleveite Gas, 297 



show a strongly marked variation between the 2nd and 3rd 

 points. Several explanations of this singularity have suggested 

 themselves to me, but I do not consider them to come within 

 the scope of this paper. I hope to make further experiments 

 bearing on this matter, and to obtain more data on the subject. 

 The results of the experiments of Jan. 31 to March 7 are 

 chiefly interesting as showing the close concordance of tem- 

 peratures registered by wires of such different thermal 

 capacity, especially when it is remembered that the source of 

 heat used was a very variable one, the pressures, as shown on 

 the indicator cards, varying from day to day, and also during 

 the day, owing to the varying gas-pressure and quality. 



1 would also point out that a difference of 100 degrees 

 shown on the thermometers corresponds to a distance of about 

 -Jj inch on the indicator cards. 



In conclusion I must express my thanks to Messrs. Johnson 

 and Matthey for the trouble they have taken in drawing down 

 the wires, several of which were of special diameters. 1 should 

 have been quite unable to carry out costly experiments of this 

 kind had it not been for the liberality of my brother, Mr. H. 

 R.J. Burstall, who has provided me with the necessary funds. 

 The apparatus has been designed and carried out in conjunc- 

 tion with him ; and I wish to express my sincere indebtedness 

 to him for his many suggestions, which have made these 

 experiments possible. 



XXIV. On the Constituents of Cleveite Gas. 

 By Professors C. Runge and F. Paschen, of Hannover* ' . 



SINCE our communication about the Spectrum of Helium f, 

 we have succeeded in constructing materially better 

 Geissler tubes, which show only traces of impurities, and 

 which emit the light of the gas itself with the greatest bright- 

 ness. With these new tubes it has not only been possible to 

 explore the infra-red portion of the spectrum as far as 10 ft 

 with the bolometer, and to discover both of the first members 

 of the two series which, according to our formula, should 

 have wave-lengths of about I'll /a and 2*03 /x, and which 

 have been found at the positions 1*120//, and 2*040 fi, but, 

 besides, the connexion of the other lines of the spectrum comes 

 out with full distinctness. 



* Translated by Dr. G. Johnstone Stoney, F.R.S., from the Proceedings 

 of the Berlin Academy of July 11, 1895. 



t Proceedings of the Berlin Academy of June 20, 1895. 



Phil. Mag. S. 5. Vol. 40. No, 244. Sept. 1895. X 



