August 1 6. 



1888] 



NA TURE 



365 



0'02I, but this is only a rough approximation. The co-ordinates 

 of the Observatory are — 



28 -6s. E. 

 52° 12' 10" N. 



A. C. Crommelix. 

 Trinity College, Cambridge, August 10. 



Macclesfield Observations. 



Many years ago, in studying Rigaud's "Bradley," I was 

 impressed by several references to extensive series of observa- 

 tions with transit and quadrant made at the observatory of 

 Shirbourn Castle, some of which Bradley evidently thought 

 worthy of comparison with his own inaccuracy. It has often 

 occurred to me that these observations, if the records still exist, 

 may well be worthy of as thorough a reduction as has been 

 given to those of other early astronomers. Perhaps some of 

 your readers can tell us something about these records of 

 1 739-89. Cleveland Abbe. 



Washington, July 30. 



A Lunar Rainbow. 



Wet Mountain Valley in Colorado is situated some 8000 

 feet above the sea, and is surrounded by mountains, the Sangre 

 de Cristo Range, on the western side, rising to some 14,000 feet 

 in its highest peaks. For the last few days we have had a 

 succession of thunderstorms — dark clouds pouring forth abundant 

 rain — which have mostly swept along the range, leaving the 

 valley clear, and often in sunshine. Last night, at 9 p.m., there 

 passed just such a storm, while the full moon shone brightly from 

 the east, where it had just risen. The result was a lunar rainbow 

 — part only of the arc, a distinct band of light, in which the 

 several colours were hardly to be observed. The phenomenon, 

 which was new to me and must surely be rare, lasted only about 

 a quarter of an hour, when the storm passed on. 



West Cliff, Colorado, July 25. T. D. A. Cockerei.l. 



GLOBULAR STAR CLUSTERS. 



T3HYSICAL aggregations of stars may be broadly 

 -*■ divided into "globular" and "irregular" clusters. 

 Although, as might have been expected, the line of 

 demarcation between the two classes is by no means 

 sharply drawn, each has its own marked peculiarities. 

 We shall limit our attention, in the present article, to the 

 first kind. 



The particles of a drop of water are not in more 

 obvious mutual dependence than the components of these 

 objects — " the most magnificent," in the elder Herschel's 

 opinion, " that can be seen in the heavens." Were there 

 only one such collection in the universe, the probability of 

 its separate organization might be reckoned " infinitely 

 infinite : ' ; and no less than one hundred and eleven 

 globular clusters were enumerated by Sir John Herschel 

 in 1864. It does not, however, follow that the systems 

 thus constituted are of a permanent or stable character; 

 the configuration of most of them, in fact, points to an 

 opposite conclusion. 



There may, of course, be an indefinite number of 

 arrangements by which the dynamical equilibrium of 

 a " ball of stars " could be secured ; there is only one 

 which the present resources of analysis enable us dis- 

 tinctly to conceive. This was adverted to, many years 

 since, by Sir John Herschel. Equal revolving masses, 

 uniformly distributed throughout a spherical space, 

 would, he showed, be acted upon by a force varying 

 directly as the distance from the centre. The ellipses 

 described under its influence would then all have an 

 identical period ; whatever their eccentricities, in what- 

 ever planes they lay, in whatever direction they were 

 traversed, each would remain invariable ; and the 

 harmony of a system, in which no perturbations could 

 possibly arise, should remain unbroken for ever : pro- 

 vided only that the size of the circulating bodies, and the 

 range of their immediate and intense attractions, were 



insignificant compared with the spatial intervals separ- 

 ating them (" Outlines of Astronomy," 9th ed., p. 636). 



But this state of nice adjustment is a mere theoretical 

 possibility. There is no sign that it has an actual existence 

 in Nature. The stipulations, upon compliance with which 

 its realization strictly depends, are certainly disregarded 

 in all stellar groups with which we have any close acquaint- 

 ance. The components of these are neither equal, nor 

 equally distributed. Central compression, more marked 

 than that due simply to the growth in depth inward of 

 the star strata penetrated by the line of sight, is the rule 

 in globular clusters. The beautiful white and rose-tinted 

 one in Toucan shows three distinct stages of condensa- 

 tion ; real crowding intensifies the " blaze " in the 

 middle of the superb group between tj and £ Herculis ; 

 in other cases, the presence of what might be called a 

 nuclear mass of stars is apparent. Here, then, the "law 

 of inverse squares" must enter into competition with the 

 " direct " law of attraction, producing results of extra- 

 ordinary intricacy, and giving rise to problems in celestial 

 mechanics with which no calculus yet invented can 

 pretend to grapple. 



Sir John Herschel allowed the extreme difficulty of even 

 imagining the " conditions of conservation of such a system 

 as that of to Centauri or 47 Toucani,&c, without admitting 

 repulsive forces on the one hand, or an interposed medium 

 on the other, to keep the stars asunder " (" Cape Obser- 

 vations," p. 139). The establishment, however, in such 

 aggregations of a " statical equilibrium," by means of this 

 " interposed medium," is assuredly chimerical. The 

 hypothesis of their rotation en bloc is countenanced by no 

 circumstance connected with them. It is decisively 

 negatived by their irregularities of figure. These objects 

 are far from possessing the sharp contours of bodies 

 whirling round an axis. Their streaming edges betray a 

 totally different mode of organization. 



Globular clusters commonly present a radiated appear- 

 ance in their exterior parts. They seem to throw abroad 

 feelers into space. Medusa-like, they are covered with 

 tentacular appendages. The great cluster in Hercules 

 is not singular in the display of "hairy-looking, curvi- 

 linear" branches. That in Canes Venatici (M 3) has 

 " rays running out on every side " from a central blaze, in 

 which " several small dark holes " were disclosed by Lord 

 Rosse's powerful reflectors (Trans. Roy. Dublin Society, 

 vol. ii. p. 132, 1880) ; showing pretty plainly that the 

 spiral tendency, visible in the outer regions, penetrates in 

 reality to the very heart of the system. From a well- 

 known cluster in Aquarius (M 2), " streams of stars 

 branch out, taking the direction of tangents" (Lord Rosse, 

 loc. cit. p. 162). That in Ophiuchus (M 12) has stragglers 

 in long lines and branches, noticed by the late Lord 

 Rosse to possess a " slightly spiral arrangement." 

 Herschel and Baily described a remarkable group in 

 Coma Berenices (M 53) as "a fine compressed cluster 

 with curved appendages like the short claws of a crab 

 running out from the main body" (Phil. Trans., vol. cxxiii. 

 p. 458). 



We find it difficult to conceive the existence of " streams 

 of stars" that are not flowing; and accordingly the per- 

 sistent radial alignment of the components of clusters 

 inevitably suggests the advance of change, whether in the 

 direction of concentration or of diffusion. Either the tide 

 of movement is setting inward, and the " clustering power " 

 (to use Sir William Herschel's phrase) is still exerting 

 itself to collect stars from surrounding space ; or else a 

 centrifugal impulse predominates, by which full-grown 

 orbs are driven from the nursery of suns in which they 

 were reared, to seek their separate fortunes, and lead an 

 independent existence elsewhere. It would be a childish 

 waste of time to attempt at present to arrive at any 

 definite conclusion on so recondite a point ; but if the 

 appeal to " final causes" be in any degree admissible, it 

 may be pointed out that mere blank destruction and the 



