492 



Profs. Liveing and Dewar. [June 10, 



for quotation instead of the maturer product of his latest study. This 

 kind of selective quotation does not represent the true scientific spirit. 



The quotation from Dr. Watts' paper ("Phil. Mag.," 1869) which 

 Mr. Lockyer prefaces by this remark — " His work was thus sum- 

 marised by himself," — has in reality no reference to a complete summa- 

 tion of his experiments, but has solely to do with " the typical form of 

 the first carbon spectrum, that obtained when olefiant gas and oxygen 

 are burnt together in an oxyhydrogen blowpipe jet." That this is the 

 correct view of Dr. Watts' position, the following extracts will show : — 

 " The spectrum obtained from cyanogen varies with the mode of pro- 

 duction. The flame of cyanogen in oxygen exhibits 7, 8, and e ; the 

 red group is replaced by a series of bands which show an opposite 

 character to the rest of the spectrum, inasmuch as each band is 

 brightest at the most refracted edge. If cyanogen be burnt in air in- 

 stead of in oxygen these bands are more numerous, extending nearly 

 to 8 and replacing 7, which is then not to be seen. Instead of the 

 group / we have two very brilliant groups of lines, which includes 

 seven lines, and 0, which is composed of six lines." 



The two very brilliant groups of lines which are referred to as £ and 

 6 by Watts, and which distinguish the flame of cyanogen, are two of 

 the sets of channelled groupings which we, in our paper on " The 

 Spectra of the Compounds of Carbon with Hydrogen and Nitrogen," 

 referred to a compound of carbon with nitrogen. Further on, in the same 

 paper of Dr. Watts, the following occurs : — " In comparing the spectra 

 of fig. 1, we notice that the changes take place at the ends of the spectra, 

 the central groups, 7, 8, c, remain substantially the same. If we pass 

 from the spectrum of the olefiant gas-flame to that of the cyanogen- 

 flame, we find the change at the blue end of the spectrum consisting 

 in the disappearance of the group / and its replacement by the groups 

 £ and 0. The group / is not absolutely proved to belong to carbon 

 (that is, it may be caused by carbonic oxide or carbonic anhydride) ; 

 but the groups £ and 0, since they are common to carbonic oxide, 

 cyanogen, and naphthaline, must be due to carbon, and their presence 

 may, with much probability, be attributed to the higher temperature 

 of the cyanogen-flame." * 



* When Mr. Lockyer makes the following statement :— " I have also repeated 

 Morren's experiment and confirmed it, and I have also found that the undoubted 

 spectrum of cyanogen is visible neither in the electric arc nor in the surrounding 

 flame," — he is referring to a different part of the spectrum altogether from the one 

 we have been discussing, and the application of the term " undoubted " to the 

 specific part of the spectrum to which he here refers is simply an expression of 

 his own view. Angstrom and Thalen in their " Recherches sur les Metalloides," 

 1875, have the following passage : — " Les groupes nuances de cyanogene, situes dans 

 les parties bleues et violettes du spectre, se montrent aussi, soit quand l'etineelle 

 traverse la partie luisante d'une flamme a. gaz, soit dans l'arc volta'ique entre des 

 electrodes du charbon d'une pile puissante. Cependant, ce dernier spectre du 



