136 



KNOWLEDGE 



[June 1, 1895, 



mann and by Col. Markwick is at least not inconsistent with 

 them, and the latter observer is ready to accept a slight 

 variation in the light of Mira near maximum. — E.W.M.] 



Sir, — During the month of February, 1895, I could 

 only observe a steady increase in the light of ]Mira Ceti. 

 The fluctuations observed by Mr. David Flanery can be 

 explained, I venture to think, by errors of observation ; 

 especially from the influence of twilight and of the full 

 moon of February Oth. When I say that the light of 

 Mira was on the increase throughout the month of February, 

 I mean up to February 15th or 20th, when the maximum 

 was reached. Yours very truly, 



•T. PL.^SSMiNN. 



Dear Sir, — I can quite endorse the statement of Mr. 

 Flanery as to fluctuations of light of some of the long 

 period variables at maximum. I think there can be no 

 doubt that at or near the maximum the light varies a little ; 

 in the case of Mira, hardly enough, however, to enable the 

 deduction to be drawn that tu-o maxima had occurred. 

 The curve of brightness of Mira, for season 1886-7, as 

 delineated on p. 303, Yol. lY., Journal <//' the British 

 Astronomical Asxoaation, will show what I mfan. The 

 general "mountain," or "hump," is broken by several 

 small hillocks or irregularities which are hardly high 

 enough to be called separate maxima. This season my 

 observations of Mira indicate the same phenomenon. The 

 day of actual greatest brightness was March 12th. Of 

 course, too much stress should not be laid on small 

 differences of brightness, for they may not always be real, 

 but be due to errors of judgment in estimating brightness, 

 to the varying brightness of the background on which the 

 star is seen, and last, but not least, to the notorious 

 difliculty of estimating correctly a ruddy star, such as Mu-a 

 is at maximum. 



I have noticed the same feature in U Orionis (see p. 

 531, Yol. LIV., M<inthhi Xotires, B.A.S.), and think 

 that, after making allowance for all possible sources of 

 error, there is a real fluctuation, in the case of the above- 

 mentioned stars, similar to what is seen in the curve of 

 sunspot frequency. Yours sincerely, 



Ordnance House, Gibraltar, E. E. Markwick. 



April 11th, 1895. 



Deak Sir, — I am sorry not to be able to say anything 

 for certain about Mr. Flauery's observations, as neither of 

 his stars are on my working list, .ludging, however, by 

 analogy, I should require very strong evidence before I 

 placed much confidence in the observations. 



In my opinion very little faith can be put in visual 

 observations to less than a magnitude, unless several 

 comparisons are made with stars of which the magnitudes 

 have been accurately estimated beforehand. In my own 

 observations, I always compare with five stars, of which I 

 have estimated the magnitudes over a period of at least 

 six months, taking the mean value as my result. 



Yours very truly, 



25, Bryanston Square, C. E. Peek. 



AprU 10th, 1895. 



[Mr. Flanery writes again, under date April 11th, to 

 point out that Pi Leonis appeared to him to reach its 

 maximum on March 22nd-23rd, twenty-two days ahead of 

 the time predicted by Mr. Sadler (Knowledge, for April, 

 p. 91), and that this is the second year that this has 

 happened. Mira, on the other hand, was many weeks 

 behind time, fluctuated greatly whilst under observation, 

 and at its brightest did not reach much more than 4-5 

 magnitude.] 



PRISMATIC COLOURS OX THE CLOUDS. 

 To the Editor of Knowledge. 



Sir, — I have recently observed on several occasions the 

 phenomenon of prismatic colours on the clouds, caused by 

 the sun shining through a thin veil of cirrus or cirro- 

 cumulus. Sir J\ Herschel (" Lecture on Optics ") alludes 

 to this as an "extremely rare phenomenon," which is 

 doubtless true as regards England, but in this clear at- 

 mosphere I have seen it four times during the last two 

 mouths, though I did not remark it during the earlier part 

 of the year. 



The colours are not vivid, but may be described as 

 beautiful transparent flushes, like those on the haliotis and 

 other well-known ornamental shells. This morning they 

 were more perfect than I have seen them before, and 

 formed a double circle round the sun on a veil of light 

 cirro-cumuli, passing at a great rate, the tinge next the 

 disc being opalescent white, then rose-pink, and beyond a 

 greenish-blue, the two latter being repeated, after an 

 interval, on the second ring. At one point all the 

 prismatic colours were visible, but less brilliant, of course, 

 than in the spectrum. Beneath, near the horizon, was a 

 long, wavy line of milky-white cloud, of which the brightness 

 varied slightly, brighter patches seeming to shift along it. 



The parhelia of Arctic voyagers are represented here by 

 the not infrequent appearance of irregular, faintly prismatic 

 patches (" sun-dogs ") on each side of the sun when low. I 

 have only once or twice seen two, and then the right hand 

 patch was much the fainter. Commonly only the one on 

 the left of the spectator is %-isible. The phenomenon of 

 " moon-dogs " is much rarer. On the one occasion that I 

 have seen it, only oi)e was visible, and that was colourless. 



It is not uncommon to see auroral appearances in the 

 daytime. On the afternoon of October 30th, between 

 three and five, I saw an unusually extensive display, the 

 sun being about twenty degrees above the horizon. An 

 arch of a smoky-brown colour advanced and raised itself 

 nearly to the altitude of the sun. At one time it formed 

 quite a broad band, then grew thmner, varying again more 

 than once. Below it were several broken bands of the 

 same colour, fringed with opalescent white, and below 

 them a darker strip of a dull coppery purple. Early in 

 the phenomenon I saw a milky hue, in a thin irregular 

 band, spread itself to some distance over the clear sky, 

 something like a streamer in the nocturnal displays, only 

 horizontally instead of vertically. On the horizon were 

 small irregular bands of a coppery colour, but displaying 

 faintly prismatic tints. At one side of the large arch a 

 small strip of sky was of a beautiful green ; a green tint 

 was also observable in the opposite quarter of the horizon. 

 Perhaps this was due to the contrast between the blue and 

 the dun colour of the arches, or perhaps to the spreading 

 of a strip of opalescent white over clear sky. 



The appearance of an auroral cloud at the same time as 

 the prismatic colours in the first-described phenomenon 

 seems a curious coincidence, if nothing more. I should be 

 grateful for any further information on the point, and also 

 with regard to the so-called " sun-dogs." 



Ebor, Manitoba. Arthur Paery, B.A. 



[The " phenomenon " of the prismatic colours on the 

 clouds may be observed frequently when nearing the snow 

 line on the Swiss Alps.] 



Note. — The first part of this description seems to refer 

 to corona rather than halo ; corona being generated by 

 water refraction, halo by that of ice. Yery beautiful 

 colours do occur, though rarely, in corona? ; but prismatic 

 colours in clouds are not a great rarity, excepting in a very 

 brilliant form. There ought always to be some colour 



