October 23, 1914] 



SCIENCE 



595 



a female in general appearance as its brood 

 sisters. Several skilled poultrymen, wten 

 shown the bird, have unhesitatingly pro- 

 nounced it a pullet. 



Aside from a perfectly clear record, the 

 marks of the operation, which are still visible, 

 show that the bird when operated on must 

 have been a male. 



While possible that this particular indi- 

 vidual may owe its feminized character to a 

 constitutional condition, such as hen feather- 

 ing, such an assumption is extremely improb- 

 able. Eather, it seems more probable that the 

 bird has actually been feminized by the im- 

 planted ovaries in similar fashion to the rats 

 and guinea-pigs of Steinach. 



A. full account of the bird will be published 

 after it has been under observation for several 

 months. 



h. d. goodaie 



Massachusetts Experiment Station, 

 Amherst, Mass. 



a third order rainbow 



To THE Editor of Science: On September 

 11, as I stood near the lake in Beardsley Park, 

 Bridgeport, Conn., I observed a rainbow in 

 such an unusual position that it seems to be 

 worthy of some short description. The rain- 

 bow was first noticed about a quarter of five 

 in the afternoon, with the sun perhaps 60* 

 from the zenith. The sky in general was clear, 

 though there were heavy clouds above the 

 eastern horizon and very light cloud streaks 

 between the observer and the sun, with a few 

 fleecy clouds near the zenith. 'So rain was 

 falling, and probably none had fallen in the 

 region for some time, nor was there indication 

 that any would fall for hours ; yet, between the 

 observer and the sun, some 10° from the 

 zenith, there appeared between two of the 

 clouds a distinct rainbow, clearly observed by 

 others whose attention was called to the phe- 

 nomenon. 



The bow was rather short, not over an eighth 

 of a circumference, convex toward the sun, 

 and showed plainly the usual rainbow colors. 

 Not until the bow had faded to such an ex- 

 tent that the colors were no longer marked was 



it recalled that no accurate statement of the 

 order of colors could be given. It is my im- 

 pression now that the red was on the convex 

 side. 



Wood's " Physical Optics," second edition, 

 p. 343, gives for the deviation produced by K 

 internal reflections in a sphere 



D = 2 (i — r) -I- fc(7r — 2r) 

 and for minimum deviation. 



For K=.Z, this gives 



i =76=50, 

 r = 46° 55', 

 I> = 31S° 20', 



whence the angle between the emergent and 

 incident light would be about 42°. This would 

 agree fairly well with the rough estimate of 

 50°. Hence the conclusion that the rainbow 

 observed was the result of three internal re- 

 flections within suspended drops of such small 

 size and number as to give no appearance of 

 a cloud. 



Various authorities, however, state with 

 more or less emphasis that the bows corre- 

 sponding to three reflections are never seen 

 on account of the much more intense direct 

 light from the sun. In the case cited above it 

 would seem that the light clouds directly be- 

 tween the observer and the sun served to 

 diminish the intensity of the direct light to 

 such an extent that the bow was plainly seen. 



This seems to be the only explanation for 

 the bow, but considering the very light clouds 

 noted above, the observation is all the more 

 remarkable. H. W. Farwell 



A SOLAR HALO IN VIRGINL4 



The solar halo, a sketch of which is ap- 

 pended, was visible over a considerable portion 

 of east Virginia for several hours on Sunday, 

 November 2, 1913. It was observed by the 

 writer at Fredericksburg, Virginia, at one P.M. 

 on that day. The phenomenon was of the 

 greatest brilliancy, the acessory " suns " being 

 at times almost as brilliant as the sun itself. 

 The great circles around the horizon were dis- 



