274 



KNOWLEDGE. 



[December, 1903. 



fourth type at 250. Some of the more notable examples 

 are indicated in the following table : — 



(Jtifarian.); 



Taking first the stars of the third type, it will be seen 

 from the examples given that there is a well-marked line 

 spectrum in addition to the flutings which form so con- 

 spicuous a feature of the visual sjiectrum. It was noted 

 by Angstrom that the flutings resembled in a general way 

 those which appear in the spectrum of manganese oxide, 

 and Lockyer in 1888 suggested that they were probably to 

 be accounted for chiefly by manganese, magnesium, and 

 lead. These identifications, however, were based upon 

 visual observations, and may need revision when more 

 precise wave-length determinations have been made. 

 Lockyer considers that some of the apparent bright 

 flutings, such as that beginning at X 6165, are real and 

 probably due to carbon, but the difficulty in distinguishing 

 between real flutings and interspaces is so great that a 

 final decision will probably not be reached until a complete 

 matching with some terrestrial spectrum is eflrected. 



The interpretation of the line spectrum is also incom- 

 plete, but much more definite so far as it goes. Taking 

 the lines generally, they bear a considerable resemblance 

 to the solar lines, but most of them are very broad, 

 and the relative intensities are so different as to greatly 

 modify the appearance of the spectrum when compared 

 with the sun. Lockyer, in 1893, announced the detection 

 in Betelgeuse, the brightest member of the group, of lines 

 due to hydrogen, iron, manganese, chromium, calcium, 

 magnesium, cobalt, strontium, and lithium, and pointed 

 out that the Hues which appear at the comparatively low 

 temperature of the osy-hydrogen flame were especially 

 strong. The blue line of calcium at K 4227, for example, 

 which is pre-eminently the low temperature line of this 

 element, is very dark and distended in the photograph 

 of a Herculis (Fig. 13). A recent investigation of the 

 spectrum of Mira which has been made at the Lick Obser- 

 vatory indicates the presence of iron, vanadium, chromium, 

 calcium, aluminium, and strontium, but manganese and 

 titanium are considered doubtful. Still more recently, an 

 investigation of x t!ygni made at Potsdam by Dr. Eber- 

 hard establishes the presence in that star of calcium, 

 titanium, chromium, vanadium, iron, and magnesium. 

 Hence, combining the different results, and including 

 sodium from the very obvious D lines, we may take the 

 following list as representing the elements so far identified 

 in the third type stars : — 



Aluminium. Iron. Sodium. 



Calcium. Lithium. Strontium. 



Chromiiun. Magnesium. Titanium. 



Cobalt. Manganese. Tanadium. 

 Hydrogen. 



The identification of vanadium and titanium may possibly 

 turn out to be of special significance, on account of the 



predominance of lines of these elements among the lines 

 which are most widened in the spectra of sunspots. Dr. 

 Scheiner has recorded that some of the stellar lines are 

 shaded on one side in a manner which recalls the appear- 

 ance of some of the lines in spot spectra, and it has been 

 - suggested that the greater part of the surface of a third 

 type star may be in a condition corresponding with that 

 of a solar spot. Attempts have also been made to account 

 for the variability of so many of the third type stars by 

 supposing a fluctuation of spotted area corresponding with 

 the eleven-yearly period of the sun, but it is by no means 

 clear why many of the stars exhibit no changes of magni- 

 tude in spite of their supposed spotted surfaces. 



Passing next to the fourth type stars, we have already 

 noted that the characteristic feature of the spectrum, in 

 visual observations at least, is the presence of dark flutings 

 fading off towards the violet ; one lies in the yellow green 

 and begins at X 5635, one in the green near b, X 5165, and 

 the other in the blue just beyond F, X4737. Secchi's dis- 

 covery that these were due to carbon vapour has received 

 abundant confirmation, and besides these the Yerkes photo- 

 graphs show the violet and ultra-violet flutings of the 

 cyanogen spectrum, beginning at X4609, 4216, and 3883. 

 The lines recorded in these stars in visual observations do 

 not number a dozen, and only sodium, iron, and hydrogen 

 could be recognized. The new photographs, however, reveal 

 the existence of .hundreds of lines, and, thanks to advance 

 proofs with which Prof. Hale has been good enough to 

 furnish me, I am able to give the following list of elements 

 which have now been identified in the fourth type stars : — ■ 



Calcium. Iron. | Sodium. 



Carbon. Magnesium. Titanium. 



Chromium. Manganese. Vanadium. 



Hydrogen. Nickel. And possibly others. 



The representation of calcium by the blue line at X 4227, 

 which is of great strength, and by the H and K lines, is 

 worthy of special mention, on account of the similar 

 appearance of these important lines in stars of the third 

 type. 



Prof. Hale concludes that the carbon and metallic 

 vapours are very dense and lie immediately over the photo- 

 sphere, while above these rise other vapours or gases pro- 

 ducing bright lines to the number of about 2(.U(, none of 

 which have been identified with certainty. Though Prof. 

 Hale seems to have convinced himseK of the reality of 

 these bright lines, it is by no means impossible that 

 Lockyer may be right in considering them as mere effects 

 of contrast. 



It is interesting to find that the relatively low tempera- 

 ture which is suggested by the flutings in the fourth type 

 stars is also indicated by the discussion of the line spectrum. 

 Prof. Hale remarks that while arc and flame lines, such as 

 the calcium line 4227 are of great strength, spark lines, such 

 as that of titanium at 453414, are less prominent or entirely 

 missing, and though he adds that this does not certainly 

 prove that the temperature of the reversing layers of the 

 fourth type stars is lower than in the case of the sim, the 

 evidence is entirely in favour of this supposition. 



It has been further ascertained that many of the lines 

 which are widened in sunspots appear as strong dark Unes, 

 and the suggestion naturally is that spots similar to those 

 which appear in the sun may be very numerous in stars of 

 the fourth type, as has been previously suggested in the 

 case of stars of the third type. The variability of so 

 many stars of the fourth type is looked upon as favouring 

 this supjjosition. 



The place of the third and fourth type stars in the 

 supposed evolutionary series has been the subject of much 

 discussion. Vogel considers that after the solar stage is 

 passed, a star develops into one of the third or one of the 



