The Theory of Temporary Stars. 9 



lines together with those of helium^ Among the nebular lines, 

 /\ 4363 made its appearance on 22nd March'-^, while A5007 was 

 certainly, and X3869 possibly, observed on 28th March, the two 

 latter showing themselves as extensions towards the violet of 

 /\5019 and H^; on this last point all the observers are in 

 agreement. 



By 19th June, owing to the gradual dying out of old lines and 

 development of new ones, the bright-line spectrum had closely 

 approximated to the nebular type^ ; the periodic variations had 

 meanwhile largely died out, though the steady fall in brightness^ 

 still went on. 



The apparent transformation of the star into a nebula was 

 not a surprise, as the same fate had befallen Nova Aiirigae ; 

 owing, however, to the faintness of that object, the process in its 

 case could not be watched, and was only ascertained to have 

 taken place several months afterwards, when the nova, for some 

 unknown reason, experienced a great revival in its light. Nova 

 Cygni is also known to have undergone a similar change, but in 

 its case — though the measurements were not precise enough to 

 allow any dogmatising on the subject — it seems likely that the 

 two principal nebular lines were present in the spectrum from 

 the first, which in the early spectrum of the other two they 

 certainly were not.' It may be added that, as far as the scanty 

 observations go. Nova Aquilae^' appears to have had a similar 

 history to Novae Aicrigae and Fersei, a history which we may, 

 therefore, regard as typical of novae in general. 



By the beginning of August the bright-line spectrum of Nova 

 Persei had fully assumed the nebular type, so far as the wave- 

 lengths of the lines were concerned ; moreover, five nebular lines 

 of unknown origin (among which three, forming a group, of 

 wavelengths 3869, 3968^ and 4363, are characteristic nebular 



1 Lockyer, M.N., Ixi., App., p. 37. 



2 Sidgreaves, loe. cit. 



3 Pickering, Ap. J., xiv., p. 81. 



4 Carefully recorded by Gore, M.N., Ixii., p. l.')6. 



5 The spectrum of Nova Cj/ijni afterwards reverted to the stellar type (Harvard Obser- 

 vatory Annual Report, 1879-80, p. 7) just as those of Xovae Cassiopeiae and Caronae appear 

 to have done. Lockyer asserts (M.N., Ixi., App., p. 19) that Xova Cygni. is now a nebula. 

 What his authority for the statement is I do not know. 



6 Pickering, Ap. J., xiii., p. 172. 



7 The Lick observers show conclusively that these must not be identified, as the}' have 

 been in the past, with helium and hydrogen lines. 



