456 Mr. James C. M c Connel on the 



(16) " October 4. Temp. 43°'0. Lunar fog-bow at 23 h> 

 Kadius to inner edge (about) 38° 5 / . 



(17) " October 5. Temp. 36°'l. At 2 h. fog was begin- 

 ning to blow across the hill-top, and on it a distinct lunar 

 fog-bow was seen with traces of a faint second bow outside it. 

 The following measurements of the inner bow were got : — 



O / 



Radius to inside edge ... 35 4 

 „ outside edge . . . 41 



" Temp. 34 0, 3. A similar lunar fog-bow was seen at 3 h. ; 

 there appeared to be a faint trace of red about the outer edge 

 of the inner bow. 



" Temp. 34 0, 1. The fog-bow was seen again at 4 h. and 5 h. 

 The following measurements were made at 4 h. : — 



o / 

 Radius to inside of inner bow . . 36 3 



„ outside of inner bow . . 41 



(18) " October 15. Temp. 25°'9. At 14 h. double fog- 

 bow and glories observed ; no measurements got." 



A remarkable feature of this list is the number of double 

 fog-bows. Out of eighteen bows no fewer than ten were 

 double. In one case (17) the outer bow was probably a form 

 of the secondary rainbow ; for the inner bow, according to the 

 measurements, extended at one edge almost up to the normal 

 radius of the primary rainbow. But in the other case, in 

 which measurements were taken (7), both bows were within 

 this radius, and in (3), (6), (7), (10), and (13) the inner bow 

 had the red inside. Assuming that the two bows are the 

 primary bow and its first supernumerary, the reversal of the 

 normal order of the colours in the inner bow is a direct result 

 of Airy's theory, and my main object in writing the present 

 article is to point this out. 



The divergence of the principal bow from its normal 

 position, and of the first supernumerary from the principal 

 bow, depends on the ratio of the radius of the drops to the 

 wave-length of the light. As it is greater with a given kind 

 of light the smaller the drops, so with given drops it is greater 

 the greater the wave-length. For red light it is greater than 

 for blue. As long as it is small for both, it only brings the 

 red bow rather nearer the blue ; but, when it is greater, it 

 may bring the red bow to coincide with the blue ; and it may 

 become so great — with the first supernumerary at any rate — 

 as to put the red bow well within the blue. This is the essence 

 of the explanation ; but, to treat the matter properly, a fuller 

 statement of Airy's results is required. 



