110 



KNOWLEDGE. 



[JiNE 1, 1898. 



very small indeed, or very dim compared with our sun ; 

 and assuming them to have a brightness equal to the 

 brightness of the sun's photosphere, they would not much 

 exceed our earth in size. 



When we look at the arrangement of the stars in this 

 cluster, we find that we have evidence of the association 

 of many of the stars into groups linked together by faint 

 nebulous bands. Long ago Sir John Herschel noticed 

 the curvilinear streams of stars radiating from the central 

 regions of clusters, and, as mentioned in the last number, 

 he remarked on the ditiiculties of conceiving of the per- 

 manent conservation of such a system as a globular 

 cluster, if the individual stars are conceived of as moving 

 round a common centre of gravity. Miss Gierke, in her 

 " System of the Stars," remarks on the " radial alignment 

 of the components of clusters inevitably suggesting the 

 advance of change, whether in the direction of concen- 

 tration or diffusion ; either the tide of movement is setting 

 inward, and the ' clustering power ' (to use a favourite 

 phrase of Sir WiUiam Herschel's) is still exerting itself to 

 select stars from surrounding space, or else a centrifugal 

 impulse predominates."-' 



It seems to me that stars falling towards the cluster 

 from surrounding space would probably not be arranged in 

 rows, and that the forms of several of the radiating groups 

 of stars which are seen to be linked together by bands of 

 nebulosity indicate that the group is moving from the 

 central parts of the cluster outwards, through a resisting 

 medium which is dense enough to contort the outward 

 moving mass. Possibly the stars may fall back singly, or 

 the stars and nebulous matter may fall back again to the 

 cluster in a cooled and non-luminous form, but I do not 

 trace any evidence of prominence forms with their heads 

 turned towards the centre of the cluster, or any linear 

 groups of stars forking towards the centre of the cluster. 

 Though there are many such linear groups of stars wliich 

 branch or fork outwards, the physical interpretation of 

 these prominence-like and branching structures seems to 

 me so important that I have made an effort to exhibit 

 them to the readers of Knowledge. I am indebted to Mr. 

 McClean, of Tunbridge Wells, for the excellent enlarge- 

 ments of the photographs of this cluster made at the 

 Lick Observatory, which are sho\™ in the collotype plate, 

 as well as for the loan of a contact copy from one of the 

 Lick negatives, a close study of which I have found very 

 instructive. The blocks have been made from enlarge- 

 ments of a photograph of the cluster taken by the Brothers 

 Henry at the Paris Observatory, and they are quite un- 

 touched, though the process has been arranged so as to 



give prominence to the faint 

 ligatures of nebulosity joining 

 the lines of stars. 



The prominence-like structure 

 shown m Fig. H should be com- 

 pared with the collotype plate 

 from the Lick photographs, as 

 well as with the cluster as shown 

 in Fig. 1. It seems to be a broad- 

 headed structure, with a con- 

 torted, spirally - twisted stem 

 joining it with the central mass 

 of the cluster. In the right- 

 hand picture on the plate made 

 from the Lick photograph of 

 July 28th, 1891, the narrow 

 stem beaded with stars can be traced, while in the 

 denser picture of May 24th, 1892, the bright head of 



Fig. 3. — Promineiice-uke 

 Structure. 



FitJ. 4. — Strings of Stars linked 

 together by Nebulosity. 



* " The System of the Stars," p. 244. 



the prominence seems to be almost severed from the 

 curving lower part of the stem ; but an examination of 

 all the photographs convinces me that the spreading head 

 is joined on to the lower part of the stem by a ligature 

 of nebulosity which probably takes a bold spiral sweep, 

 and joins up with the lower part of the stem at the cusp 

 marked by the bright star. On the denser Lick photograph 

 of May 24th, 1S92, at 

 least two faint strings of 

 minute stars joined by a 

 ligature of nebulosity 

 may be traced branch- 

 ing away from the 

 bright prominence-like 

 structure, or possibly 

 seen in projection in its 

 immediate neighbour- 

 hood. 



The structures shown 

 in Fig. i are from the 

 upper part of the cluster 

 to the right of the 

 prominence structure 

 referred to. The structures will be best seen by a reference 

 to the collotype plate from the denser Lick photograph, 

 but the strings of stars, and the nebulosity linking them 

 together, can also be recognized on Fig. 1, made from the 

 Henry photograph. The streams of stars are not straight, 

 and seem to afford evidence of having been contorted by a 

 resisting medium. 



Fig. 5 shows a very curious structure, which has the 

 appearance of being a spiral stream 

 of nebulous matter and stars, 

 with a spreading or forked summit. 

 This structure is immediately 

 below that shown in Fig. 4, and 

 extends radially from the right- 

 hand side of the cluster. It is 

 well worthy of close study, and 

 should be compared with the Lick 

 plates and with Fig. 1. If we 

 admit that this is a vast spiral 

 structure, it seems to follow that 

 the resisting medium through 

 which the ejected matter was pro- 

 jected upwards is sufficiently dense, 

 compared with the ejected matter, 

 to permit of vast storms causing cyclonic action. We 

 occasionally find prominence forms upon the sun, showing 

 distinct evidence of cyclonic action. A notable instance 

 of such a twisted solar prominence was photographed by 

 Colonel Tennant durmg the total eclipse of 18G8, and is 

 well shown in the plates made from his photographs 

 pubUshed in the ' ' Memoirs of the Eoyal Astronomical 

 Society," Vol. XXXVII., Plates 5o and 5i), and on a 

 larger scale in Plates G, 7, and 8 in the same volume. 

 We know that at the level of the solar prominences the 

 pressure of the solar atmosphere, or resisting medium 

 through which the prominences are projected, is so small 

 that the lines of the prominences, as shown in the spectro- 

 scope, are quite narrow, and it seems probable that the 

 cyclonic twist must be given to such prominence matter 

 at a much lower level than the level of the photosphere. 

 Similarly, it seems possible that the spiral twist shown 

 by the structures in Figs. 3 and 5 may have been given 

 within a gaseous envelope in the central region of the 

 cluster. The flattened and spreading heads of prominences 

 may be due to the rapid motion of the prominence 

 matter through a very thin resisting medium, or even 



Fig. 5. — Spiral SUvaiii 

 of Nebulous Matter iincl 

 Stars. 



