January 23, 1913] 



NATURE 



:}/ 



radius r of the cloud drops from the angular radius 

 of a ring, a divergence from theory was found on 

 this, and other, occasions. Theoretically all the rings 

 should give the same value for r, but the calculated 

 value of r was found to diminish with the order of the 

 ring outwards. 



The Japanese observers are stated in the note to 

 attribute the halo to reflected light from sun-images 

 formed on the green blades by rays refracted through 

 dewdrops. In the case of the cloud observations it 

 seems necessar}' to assume reflection from portions of 

 the clou.d itself. 



The difficult point, however, is to explain why the 

 light thus reflected should be maximum in the direc- 

 tion of the sun, or, what is the same thing, in the 

 direction of the observer. The fact that the ring 

 surrounds the shadow of the observer's head seems to 

 render such an assumption necessary. A single drop, 

 as Prof. Richarz points out, does not give maximum 

 intensity of reflection in the direction of the incident 

 light. Dr. Richarz's explanation why the cloud as 

 a whole should do so is simple and ingenious, and is 

 applicable whether the sun's rays fall normally or 

 obliquely to the surface of the cloud. Direct light 

 only penetrates into the cloud (or assemblage of drops) 

 when it finds a clear path, for if it strikes any drops 

 on the way it will be scattered or diverted by refrac- 

 tion and reflection at their walls. If light which has 

 so penetrated should then fall on the surface of a 

 drop in the interior, it will be reflected in various 

 directions, but only that portion of the reflected beam 

 which returns the same way it came can find a clear 

 path out again. Portions of the beam reflected in 

 other directions will generally find their way blocked 

 by intervening drops and be scattered. Hence the 

 inlensity of the reflected light will be maximum in 

 the direction of the source of light, and the intensity 

 will fall off rapidly with departure from that direction. 

 The observer's head (or the balloon) cuts off' the central 

 portion of the sheaf of rays which he would see most 

 brightly reflected, leaving only the peripheral portion 

 visible. 



To digress, I have a vivid recollection of one very 

 foggy winter evening when I was wintering in a 

 cottage on a wild part of the Cornish coast. Chancing 

 to throw open the casement window of the sitting- 

 room, I was for the moment quite taken aback 

 to find myself confronted by a tall sinister figure 

 looming up before the window. It was my own 

 shadow thrown on the fog by a lamp left unshaded 

 on a table in the room. 



Perhaps I may take this opportunity to record 

 another little optical observation of different character. 

 Once — I think it was towards the close of the hot 

 summer of 1908 — watching, from the top of a cliff 

 some 800 ft. high, the sun setting over the sea, I 

 saw the upper half of the disc look like a double 

 staircase ; there were three or four distinct, almost 

 rectangular, steps cut out of the limb symmetrically on 

 either side. When most of the disc had sunk out of 

 sight, the small portion remaining was suggestive 

 of the lid of a teapot with a knob on top. Some 

 lines of light cloud about the horizon showed the 

 existence of horizontal stratification in the atmosphere, 

 and the strange distortion of the solar limb was 

 evidentlv due to refraction through horizontal strata 

 with extraordinary sharpness of boundary and differ- 

 ence of density. .'Vlice Everett. 



Milbourne Lane, Esher, January 6. 



" Rosa stellata." 



In i8q8 Prof. E. O. Wooton described a remark- 

 able new rose from southern New Mexico, giving 



NO. 2256, VOL. 90] 



it the name Kosa stellala on account of the stellate 

 trichomes. The peculiar, mostly trifoliolate leaves, 

 the leaflets with cuneiform bases and more or less 

 truncate, sharply toothed apices, gave the plant an 

 unusual appearance ; while even the flowers, described 

 as ■' large and showy . . . deep rose-purple," were 

 not at all like those of the ordinary wild roses of the 

 Rocky Mountains. Through the kindness of my 

 friend, Prof. Fabian Garcia, I obtained some living 

 plants of R. stellata from the original locality in the 

 Organ Mountains. Some of these were sent to Dr. 

 A. R. Wallace, who has grown them in England 

 successfully ; the others have been growing in 

 Boulder, Colorado. Last year the plants in my 

 garden grew exceedingly well, and were most attrac- 

 tive. Certainly if R. stellata can be generally used 

 in gardens, it will be a valuable addition to horticul- 

 ture, but it probably will do its best only in relatively 

 dry climates. My wife attempted crosses with several 

 other roses, and in one case was successful in getting 

 good seed ; what will result remains to be seen. 



The fruit of R. stellata, as indicated by Wooton, 

 is large, beset with strong slender prickles. Quite 

 unlike the usual types of rose fruits, its walls are 

 dense, not at all fleshy or brilliantly coloured, but 

 corkv. The orifice is very broad, with a diameter of 

 8 mm. The bright chestnut-red seeds, about 4 mm. 

 long, are long-oval, not compressed, and therefore not 

 at all angular. All this differs conspicuously from the 

 fruit of typical Rosa. 



R. stellata, however, is not the only plant of this 

 type. Years before, Engelmann described R. minuti- 

 folia from Lower California, a plant with the same 

 general characters. In recent times, Dr. Greene has 

 separated part of Wooton's R. stellata as R. mirifica, 

 and has added a fourth species, R. vernonii. Thus 

 we have a compact group, which should, I think, 

 form a distinct subgenus or genus Hesperhodos, with 

 stellata as the type. All the species are of extremely 

 restricted distribution, which may probably be ex- 

 plained by the fact that the fruits are not adapted to 

 be eaten by birds. 



The wide-open prickly fruit suggests that this may 

 be a primitive form, as compared with true Rosa ; 

 but it is to be noted that the roses found fossil in 

 the Miocene beds of Florissant, Colorado, belong to 

 the true genus Rosa, not at all to Hesperhodos. 



T. D. A. COCKERELL. 



Boulder, Colorado, December 30, 1912. 



A Lens or a Burning Glass? 



In the latest edition of Carpenter on the microscope 

 at p. 1 19 occurs the following, evidently from the pen 

 of the late Dr. Dallinger :—" There is in the British 

 Museum a remarkable piece of rock crystal, which is 

 oval in shape and ground to a plano-convex form, 

 which was found by Mr. Layard during the excava- 

 tions of Sargon's Palace at Nimroud, and which Sir 

 David Brewster believed was a lens designed for the 

 purpose of magnifying. If this could be established 

 it would, of course, be of great interest, for it has 

 been found 'possible to fix the date of its production 

 with great probability as not later than 721-705 B.C. 

 . . . we spent some hours in the careful examination 

 of this piece of worked rock crystal, which, by the 

 courtesy of the officials, we were permitted to photo- 

 graph in various positions, and we are convinced that 

 its lenticular character as a dioptric instrument cannot 

 be made out. There are cloudy strijB in it, which 

 would nrove fatal for optical purposes, but would be 

 even sought for if it had been intended as a decorative 

 boss; while the grinding of the 'convex' surface is 



