October 1, 1895. 



KNOWLEDGE. 



233 



sections of the lines making two square millimetres each. 

 Another method is to use a millimetre rule with a thin 

 edge, and a pair of compasses ; a circle of about ten 

 centimetres in diameter and ruled into degrees upon any 

 transparent substance would also be required for the 

 purpose of measuring position angles. Armed with these 

 simple requisites, much useful astronomical work could be 

 done upon the photographs, and if any doubtful points 

 should be met with iu the course of the investigations, 

 they could be settled by reference to the original positives 

 on glass made by enlargement from the negatives, or, if 

 necessary, by reference to the negatives themselves. 

 Arrangements are being made to place these glass positives 

 at the disposal of the British Astronomical Association, 

 and thus make them readily available whenever required. 



If the work of correlating the photographs, as herein 

 suggested, is taken in hand by competent examiners, and 

 their results are found to add to astronomical knowledge, 

 there are other photographs with intervals of seven or 

 eight years between them available for investigation, which 

 could be published as occasion might indicate ; thus a "" 

 system of astronomical research would be inaugurated, 

 that must eventually add largely to existing knowledge. 



The scale of these photographs is such that a change in 

 the position of any star with reference to the other stars, 

 if it exceeds three seconds of arc in amplitude, ought to be 

 detected by careful examination ; and, if no change of 

 such small extent as this has taken place iu eight years, it 

 will be a proof that the object is very distant from us — that 

 the stars probably belong to one system, and that we must 

 wait for some years longer before we can hope to under- 

 stand the physical laws which govern the stars constituting 

 these so-called globular clusters. However, we have now 

 before us, and fixed with imquestionable accuracy, the 

 position and relative brightness of each star in the cluster 

 and in the surrounding region of the sky, so that the 

 solution of this great problem can only be a question of 

 time, and of the repetition of the methods here described, 

 or of others of a similar character. 



observations were made with the aid of field-glasses, power 

 4, out of focus ; and all at Sunderland, except that on 

 2 mo. 24, which was at Middlesbrough. The eflect of 

 atmospheric absorption upon the brightness of Mira and 

 the comparison stars was taken into account for each 

 night's resulting magnitude. 



Dati-. ,^M.r. .?'^'*"""^' Absovi). 



Muj^uituaes. tion. 



1895, 1 mo. 23.— 9.10 5-72 1-0 



25.-9.27 5-98 1-0 



29.— 6.40 4-85 2-5 



2 mo. 11.— 0.50 4-45 3-0 



19. 7.15 4-23 2-5 ^ot a very certain obser- 

 vation, as tlie smoke iu the 

 iog may not be luiiform. 



24.. — 8.28 4-98 1*1 rossiWy some irrcijnlar 



smoke, but I do not think so. 



3 mo. 3.-7.59 4-30 1-1 Moonlight. 



Note. — The scale of each of the four photographs of 

 nebula published in the September number of Knowledge, 

 is given as 1 millimetre to 12 seconds of arc ; but the 

 photo-printers have not strictly followed the directions, 

 and consequently the scale applicable to the four prints is 

 1 millimetre to 11 seconds of arc. 



3lcttcrs. 



[The Editor does not hold liimseU' responsible for tlie opiuious or 

 statements of correspondents.] 



— ■ ♦ ■■ '■■ 



MIRA. 

 To the Editor of Knowledge. 



Dear Sir, — My observations of Mira in the early part 

 of this year may perhaps be of some interest in connection 

 with the discussions on the subject, though it will be seen 

 that the later ones are affected by some uncertainty. The 

 resulting magnitudes are derived from the Harvard 

 magnitudes of the comparison stars, but that of the ruddy 

 star i' Ceti is corrected for personal equation of colour. 

 The column headed "absorption" gives the estimated 

 relative absorptive power of the atmosphere, taking 1-0 to 

 represent a clear atmosphere. With regard to the result 

 on 3 mo. 3, it should be noted that moonlight makes ruddy 

 stars look brighter ; but ruddy stars look fainter to me 

 than they do to the Harvard observers, and the resulting 

 magnitudes are given as compared with white stars. The 



Sunderland. 



T. W. Backhouse. 



CONCENTRIC RAINBOWS. 



To the Editor of Knowledge. 



Sir, — I recently saw what I believe is rather a rare 

 phenomenon. There was a short, sharp shower of rain 

 about six o'clock iu the evening, and then the sun shone 

 brilliantly and a rainbow appeared — not a particularly 

 bright one. There was no secondary bow with inverted 

 colours, but inside the primary bow were packed, one 

 inside the other, no less than three subsidiary bows fading 

 imperceptibly into one another. I once, two years ago, 

 saw a very bright rainbow with two of these subsidiary 

 bows, and also the usual secondary bow. The theory of 

 the primary and secondary bows is simple enough, but I 

 should be glad if some of your correspondents could give 

 an explanation of the theory of these subsidiary bows. 

 I gather from what I have seen stated that, provided the 

 rain-drops are equal in size, the larger the drops the 

 greater is the number of subsidiary bows. 



Arthur Kennedy. 



Ashtead. 



[The occurrence of supernumerary bows enclosed 

 witbin the prioaary one is by no means common. I have 

 only seen these on two or three occasions in many years. 

 Once I observed no less than four bows of gradually 

 decreasing width, enclosed by and concentric with the 

 primary one, the innermost being a faint colourless line. 

 There were thus five circular bands, and between their 

 widths there was a distinctly proportionate diminution 

 towards the centre of the system. On other occasions I 

 have seen three or four bands, the inner ones being very 

 faint. 



In my experience this phenomenon synchronizes with a 

 very stormy and disturbed condition of the air. The 

 bands appeared in a higher zone of the sky than the 

 ordinary rainbow is seen in, but not so high as that of 

 the ii-hite rainbow, which at times occurs in the cirrus 

 region. I once saw one of these latter at a vast elevation 

 gradually concealed from sight by a moving drift of cloud 

 underneath it, the obscuring cloud being itself far above 

 the region of the ordinary bow. 



It would appear that these involved bands, which usually 

 show little colour, indicate the d^pth and thickness of the 

 bank of rain cloud, or rather the zone of watery particles 

 originating them. If so, they naturally give a valuable 

 hint for forecasting weather, as they point to a vast 



* One of these, wliieli was seen January .jth, 1895, was described by 



the writer iu Nature. 



