OCTOBEB, 1903.] 



KNOWLEDGE. 



231 



to l>e remarked that the well-defined fan-like rays, so well 

 seen in No. 1 photograph, could not he perceived visually 

 witli anytliiug like so great distinctness in the larger 

 reflector. 



The scale of the photographs taken on plates 2| by 2 is 

 049 inch = 1°. R. C. Johnson. 



No. 1.— Borrellj'3 Conief, 1903,'July 22d., lOh. 58iu.— llh. 3fni. 



SOUTH. 



-Borrelly's Conwt, 1903, July 2M., lOli. 3(lm.— llh. Oin. 



Silts, 

 is very 

 T have 



CURIOUS SUNSET PHENOMENON. 



TO THE EDITORS OF KNOWLEDGE. 



— Tlie phenomenon dcscrilied by Lie\it. Field, it. a., 

 familiar to me. Although I have not been in India 

 frequently seen it in South Africa. It reijuires a 



bright sun and a thinly-dispersed cirrus cloud in the spots 

 where the patches appear ; I believe about 22" either to 

 right or left of the sun. I often used to watch the pheno- 

 menon when in Natal, and it could often be seen en either 

 side of the suu. 



Looking over my note books I find I recorded the same 

 thing as seen at Gibraltar on 30th May, 1893—-' a pretty 

 coloured image of the sun to the left of that luminary." 



In Ireland, January 29th, 1890, I recorded portions of a 

 solar halo visible. To the right of the sun there was a 

 brilliant white patch of light, shaded off with prismatic 

 colours. There was another patch, but not quite so intense, 

 on the left of the sun. They evidently formed parts of a 

 circle, of which the sun was the centre, for later on, at 

 about 10 a.m., I could trace the halo above the sun joiniog 

 the coloured parts, although it (the circle) was white. 

 The patches continued liright for about a quarter of an 

 hour. This, it must be noted, was iu the morning. 



Lastly, I note that on April 29th, 1902, while crossing 

 from Penzance to the Scilly Isles, I saw at about 6 p.m., 

 a rather faint-coloured patch of light 22° (or so) to right 

 of suu. There was a tendency to lengthen, vertically, as if 

 part of a halo. Part nearest suu red, rest white. Sky to 

 left of suu clear blue ; but suddenly a scrap of " mare's- 

 tail " cloud (drifting along) showed the most brilliant 

 prismatic colouring, principally red, but also partly pure 

 white. A very pretty sight. 



It is obviously a phenomenon analagous to the rainbow, 

 but instead of a reflection followed by refraction, I presume 

 it is a case of refraction and dispersion only. What is the 

 nature of the particles of watery vapour favourable to the 

 production of these coloured patches, that is, whether in 

 form of water, ice, or otherwise, I am not physicist enough 

 to say, but from my experience I am certain that a thin 

 stratum of cirrus cloud is necessary to produce them. 

 Thev are commoner than supposed, and faint traces of 

 them can often be observed by a lover of the sky who 

 keeps his eyes open. 



Salisburv. E. E. M.\ekwick, Col. 



METHOD OF ASCERTAINING MOON'S AGE. 



TO THE EDITORS OF KNOWLEDGE. 



Sirs, — Mr. Crommeliu in his interesting paper on 

 Cycles of Eclipses tells us the Saros may be used to give 

 tlie rough position of the moon and age on dates for which 

 an almanack is not available. If only the age is required 

 there is a simpler way, which requires no almanack at all. 

 Mv. Crommelm is certainly acquainted with it although he 

 has not mentioned it. I am not aware where I got it 

 from, but I continually use it mentally when I come across 

 references to moonlight phenomena, and often find errors. 

 It is to divide the year by 19, take the remainder, multiply 

 bv 11 and divide by 30. Take the remainder again, add 

 the <lay of the month, and one for every month in the year 

 after March, the result is appro.Kimately the age of the 

 moon. Thus, to take Mr. Crommelin's two dates, what 

 was age of moon March 8th. 1899, and what was age 



Then ^'^•^•' = 99, and remainder 18. 

 19 

 18 X 11 



30 

 and we get age of moon, 2(3 days. In the second case, 



1809 „^ , . , , -t X 11 



- = 9.'), and remainder 4. ., , 

 19 <iO 



14. Add this to January 12th, and we again get age of 

 moon, 2(j days. 



J ust as au instance later in the year. What was age 



January Pith. 1809 : 



= G, and remainder 18. Add this to March 8th, 



= 1. and ivmainder 



