December, 1908.] 



KNOWLEDGE. 



275 



fourth typo according to the chemical combinatiou which 

 is first formed in its atmosphere (the bands described in 

 the foregoing as carbon bands being attributed by some 

 to hydrocarbon, and not to the element carbonj, and 

 Duner has adoptgd the same view. That the fourtli type 

 stars follow the second may be considered as fully 

 established ; for besides the evidence afforded by the line 

 spectrum, there is in the solar spectrum a well-marked 

 absorption fluting iu the ultra-violet at \ 3883, corre- 

 sponding with one of the flutings seen in the Yerkes 

 photographs of the fourth type stars, and, according to 

 Rowland, there are many faint solar lines which agree 

 with the constituent members of the carbon fluting .516.5. 



It is the place of the third type stars which remains 

 somewhat doubtful. They do not seem to come between 

 the second and fourth types, and probably few would be 

 inclined to place them after the fourth type, although the 

 carbon and cyanogen flutings typical of the fourth type 

 stars survive the temperature of the electric arc while 

 flutings of other substances fail to do so. 



Prof. Hale practically adopts Vogel's view, on the 

 ground that the resemblances between third and fourth 

 type stars are greater than the points of difference ; they 

 resemble each other, for instance, in colour, tendency to 

 variability, spectra, possible presence of sun-spots, physical 

 condition and probable relationship to solar stars, and 

 " they should therefore be classed together as co-ordinate 

 branches leading back to stars like the sun." The spectro- 

 scopic similarity of the two types, according to Prof. Hale, 

 is intensified by the presence in both of the cyanogen 

 flutings in the violet, beginning at X 4609. Apart from 

 the differences in the other flutings, the chief important 

 divergence is in the distribution of the two types over the 

 celestial sphere, the third type stars showing no preference 

 for the region of the Milky Way, while the stars of the 

 fourth type are distinctly condensed in this neighbourhood. 

 Assuming the third and fourth type stars to be alterna- 

 tive forms of decaying solar stars, the greater part of the 

 story of stellar evolution, according to the views of Vogel 

 and others, may be shortly expressed as follows : — 



Type (Orion stars). 



Type I. 



Type II. 

 Type III. -Type lY. 



Dark stars. 



There is good reason to believe that the maximum of 

 cosmical temperature is manifested in the Orion stars, and 

 it is possible that this stage is reached by the condensation 

 of nebulae through a stage represented by the Wolf-Eayet 

 stars, the spectra of which chiefly consist of bright lines. 



Lockyer finds a difficulty in believing different kinds of 

 stars to represent a similar stage of development, and, as 

 previously explained, divides the stars into two series, one 

 supposed to include stars of increasing temperature, and 

 the other the stars which are becoming cooler. Stars of 

 the third type take a low position on the rising branch 

 of his "temperature curve," following the uebulse, and those 

 of the fourth type end the series of luminous bodies on 

 the descending branch. A general statement of Lockyer's 

 classification, embodying this idea, is as follows : — 



IIlUHEST TEMPERAinRK. 



Notwithstanding the divergence of opinion on some 

 jioints, there is a general acquiescence in the view that the 

 matter composing the stars is essentially the same as that 

 with which we are acquainted on the earth. This leading 

 idea is admirably expressed by Sir William and Lady 

 Huggius in the following passage in their " Atlas of 

 Representative Stellar Spectra " : — " As the conclusion of 

 the whole matter, though there may be no reason to 

 assume that the proportions of the different kinds of 

 chemical matter are strictly the same in all stars, or that 

 the roll of chemical elements is eoriially conaplete in every 

 star, the evidence appears tq 1^6>strong that the principal 

 types of star-spectra should nof be interpreted as produced 

 by great original ditt'eren*es of cheiiiical' constitution, but 

 rather as successive ^8taggs■ of ^ evolutional progress, 

 bringing about such altered cc^ditfons o£-density, tempera- 

 ture, and mingling of ^tellar gastjs, as are sufiicient 

 presumably to account for the sjjeetral differences observed, 

 even though witk-.a^m' presebt; knowledge a complete 

 explanation maj^ot b§ fiirtiiComing." 



Investigations are stil*Tn progress in many lands, and 

 it is not too much to expect that sooner or later the story 

 of celestial evolution will be completely elucidated. 



THE SUNSPOTS OF 1903, OCTOBER. 



By E. Walter Maundee, f.k.a.s. 



In a few short months we have passed from a period of 

 extreme quiet on the sun to a period in which great spots 

 have begun to show themselves. I should wish to use 

 this teiTU of " great spot " in a strictly definite and limited 

 sense. For iu every sunspot cycle there are a few groups 

 which .stand out, beyond all others, by their pre-eminent 

 size, the giants of their kind. "Normal" spots, to use 

 Miss Brown's nomenclature, or "regular" spots, to use 

 that of the Glreeuwich Results, that is to say spots nearly 

 circular in shape and with umbra central and also nearly 

 circular, generally have an area of from 100 to 300 

 millions of square miles. Above this size they generally 

 show a much more irregular and complex structure, and 

 at times of maximum activity, groups with a total area of 

 700 — a million of square miles being taken as the unit — 

 are pretty frequent. But the groups which attain a mean 

 area of double this amount are few indeed. Only sixteen 

 were recorded from the beginning of the Greenwich record 

 right up to the end of September of the present year. 

 And if we take a higher limit still, and fix an area of 2000 

 as the smallest for a really " great group," we shall find 

 that that maximum is an unusual one that can boast of 

 half-a-dozen members of the class. Necessarily, solar 

 disturbances on this scale arc strictly confined to the years 

 of maximum activity, and the last two maxima only show 

 five such groups between them. 



The appearance at the east limb on October 4 of a 

 regular spot, the first member of a group which was fully 

 entitled to be classed amongst these giants, marks, there- 

 fore, an epoch of great interest in the progress of the 

 present cycle. The interest iu it is greater when we bear 

 in mind that in previous cycles tlie increase in activity has 

 followed so rapidly after the appearance of the first really 

 great sjiot, that the largest group of the entire cycle has 

 usually been registered within six or seven months after- 

 wards. Spot groups continue to increase iu frequency, 

 indeed, for a considerably longer time, but, as a rule, not 

 in size. If this precedent be followed, then the greatest 

 single disturbance of the present maximum may be 

 expected not later than April or May, 1904, whilst the 

 actual summit of the curve is rather to be expected to fall 

 iu 1905, probably somewhat late in the year. 



