ii6 



NATURE 



[October i, 19 14 



The experiments indicate that any individual sexual 

 female produces either only male-producing or only 

 female-producing parthenogenetic offspring, but not 

 both. Individual fertilised females were allowed to lay 

 on sleeved oak-leaves in May, 19 13 ; the galls pro- 

 duced by each were collected, and sleeved on oak 

 leaves in March of this year. On examining the flies 

 produced, I found that, with a few possible exceptions, 

 the grandchildren of any one sexual female are either 

 all males or all females. I have examined more than 

 qooo grandchildren of the twelve fertilised females 

 sleeved in May, 1913, and among these there are about 

 2 per cent, of exceptions to the rule that all the grand- 

 children of any sexual female are of the same sex. 

 The exceptions, however, do not occur in all the 

 sleeves, but in rather less than half, and I hav^e little 

 doubt that they are due to wild flies having been able 

 to lay eggs in the buds through the muslin of the 

 sleeves. The wild flies were exceptionally abundant 

 this spring, and it is difficult to prevent the buds from 

 pressing against the muslin in such a way that a fly 

 on the outside could not lay in them. I intend to test 

 this possibility next spring, but meanwhile the results 

 obtained make it certain that nearly, if not quite all, 

 the grandchildren of any sexual female are of one 

 sex, and that of the sexual females, those which have 

 male or female grandchildren are about equally 

 numerous. L. Doncaster. 



Cambridge, September 21. 



anti-kathode (Fig. 3.). Around this axis the effects are 

 increasingly asymmetric. This is shown by the varia- 

 tion of the white band (from a full circle to nearly a 

 semi-circle) within the shadow, and of a dark area 

 between it and the boundary of the image. The posi- 

 tions of maximum intensity have been determined by 

 the photographic and electroscopic methods (see 

 Nature, August 13, Rontgen Journal, July and Octo- 

 ber, 19 14), and are found to be always directed to the 

 optimum axis. The asymmetry has also been estab- 

 lished with V-apertures and squares of metal, where 

 the white band appears on one, two, or three (asym- 

 metric), or four sides (symmetry) according to the 

 relative positions (70° to 0°) to the axis of sym- 

 metry. The bulb must, of course, be rotated on the 

 spot as centre. In addition to these effects, already 

 described, we have now ascertained that the 

 "diffracted" rays are almost entirely re-diffracted by 

 a second edge. The re-diffraction (as in the case of 

 primary diffraction probably) occurs in two directions. 

 I.e. within and without the shadow. Moreover, these 

 effects are again definitely asymmetric when other 

 than optimum rays are used. For instance, a lead 

 disc perforated by apertures symmetrically arranged 

 is shielded by a solid disc of slightly larger dimensions 

 to secure that no rays except those already diffracted 

 by it can reach the second disc. The perforated disc 

 may be placed 0-7 cm. from the plate, and 19 cm. 

 from the anti-kathode, 27 cm. separating the discs. 



Fig I. —Asymmetric. Extreme X-rays 

 CGraz'.ng anti-kathode) 70° irom normal in 

 plane at right angles to that of kathode 

 rays. 



Fig. 2. — Symmetric. Optimum X-rays 15° from 

 true reflection and in plane of kathode rays, 

 i.e., 30° from normal to anti-kathode. 



Fig. 3.— Map of hemi.sphere of X-rays (lead 

 di>c) showing radial structure of 

 radiation. Reduced 4 diameters. 



Asymmetric 



Diffraction " and " Re-Diffraction " 

 X-Radiation. 



of 



Further work upon this subject, briefly described 

 in previous letters (see Nature, July i6, p. 507), has 

 shown that for any one position of the bulb and 

 object the angle of diffraction is constant. This angle, 

 however, varies with the distance of the object from 

 the source in accordance with a simple inverse sine- 

 law (sin 6 xdcm = 75), over the .wide range of 8° to 

 40°, and 50 to 10 cm. This and other well-marked 

 differences from light have been already noticed in 

 our previous papers, and a table of the sine values is 

 given in the present (October) number of the Rontgen 

 Journal. 



A map of the hemisphere of X-rays, charted by the 

 diffraction images of a lead disc, shows that there is 

 an "optimum," or axis of symmetry, lying in the plane 

 of the kathode rays and 30° from the normal to the 



NO. 2344, VOL. 94] 



These rays, being normal to the disc, and leaving j 

 the anti-kathode at any other than the optimum angle, ' 

 the images of the perforations will be seen (Figs, i j 

 and 2) to be increasingly displaced, distorted, and 

 altogether absent in the extreme position. The asym- j 

 metry is marked at opposite points, and not at the | 

 ends of rectangular axes as would be the case were 

 the phenomenon one of ordinary polarisation. I 



A black band outside the shadow is naturally only \ 

 observed when the rays are thus shielded, for other- | 

 wise it falls in the area of undiffracted direct radia- | 

 tion. 



In these experiments especial care has been taken 

 to keep the milliamperage constant between i and 

 1-5, by means of a variable resistance in the primary. 



I. G. Rankix. 

 W. F. D. Chambers. 

 90 Gordon Road, Ealing. 



