November 3. 1923] 



NATURE 



653 



ti 



I 



a definite orientation of the powder layer at B, which 

 corresponds to the orientation of the surface of 

 revolution and vice versa. In order to record the 

 lines over an extended angular region a screen D has 

 therefore to be provided with an opening, which for 

 any particular position of B uncovers only the 

 corresponding portion of C. The screen has to be 

 moved with uniform angular velocity and B has to 

 take the corresponding required positions. If we 

 all /3i and /Sg the glancing angles of incidence and 

 ^ , sin /J, AB 

 of emergence at B then ";„ o ~ BC ^ ^\-^?i'=-'^ is 



the angle of deviation. The relation between the 

 motion of the screen and of the powder layer becomes 

 the simple spectrometer relation when AB equals BC, 

 but this arrangement is not the most efficient for 

 obtaining beams of greatest specific intensity. With 

 the setting corresponding to a given resolving power 

 the time of exposure will depend on the angular 

 width of the region explored. The method is most 

 efficient for exploring small angular regions for the 

 exact measurement of a few characteristic " key " 

 lines, but owing to the gain in intensity by using wide 

 beams there is some saving of exposure also for more 

 extended surveys. 



A fuller discussion of the method and description 

 of the apparatus used will be given elsewhere. With 

 a small camera of this type, BC being 2-3 cm., a photo- 

 graph was taken of the first order reflection of CuK 

 radiation from the iii and 100 faces of nickel oxide 

 with 1-2 milliamp-hour exposure, the angular exten- 

 sion of the region recorded being about 20°. The 

 lines were less than o-i mm. wide and their centres 

 could be evaluated to 0-03 mm. When the greatest 

 possible intensity is required for tracing faint lines in 

 a narrow angular region, a powder layer of suitable 

 curvature which allows us to use beams of consider- 

 able angular width is of advantage. For quantitative 

 measurements, where the absorption under different 

 angles of incidence has to be taken into account, and 

 for exploring wider angular regions, a flat surface is 

 more suitable. By exposing it from different sides, 

 errors due to eccentricity in mounting can then be 

 eliminated. This procedure was used in the case of 

 the nickel oxide mentioned. J. Brentano. 



The Physical Laboratories, 

 The University, Manchester, 

 October 12. 



A Large Sarsen Stone. 



A SARSEN stone of unusual size, for this district, has 

 recently been found in the gravel pit belonging to the 

 Hounslow Sand and Gravel Co., and through the 

 courtesy of the manager, Mr. Ralph Wallis, I have 

 been permitted to pay several visits for purposes of 

 investigation and photography. 



In section, the pit shows : 



SoU 1 ft. 



Indurated mud, like warp i ft. 6 in. 

 Loamy gravel, penetrated by 



the warp (averages) . . 7 ft. 



" Clean " gravel and sand . 8 to 21 ft. 



resting on London Clay of unknown thickness. 



The sarsen (Fig. i) was found embedded to the depth 

 of I ft. in the London Clay with several others of 

 much smaller size — from a few lb. to about 2 cwt. — 

 and they were the only ones found there. It is 

 computed to weigh 6 or 7 tons, but owing to the 

 number of tubular cavities present, varying in length 

 from a few inches to 3 ft., and in diameter from t'ff to 

 2 in., even an approximate computation may have 

 to be considerably revised. Its maximum height, as 



now standing, is 5 ft. 7 in., maximum thickness i ft. 

 II in., and its maximum width 5 ft. 7 in. 



There are several interesting details which might 

 occupy too much space to describe here, but perhaps 

 I may be permitted to refer to the cruciform surface- 

 feature conspicuous in the photograph of the surface 

 which was uppermost when the block was in situ. It 

 is due to the fact that two of the long, tubular cavities 

 cross each other in the heart of the stone, this being 

 rendered visible through the erosive action of falling 



water, at some time or another, forming a basin- 

 shaped depression, 4 ft. in diameter and 7 in. deep, 

 which has exposed the internal structure. There is 

 little doubt but that the tubular cavities have also 

 been considerably enlarged and modified by the action 

 of running water. A few striations on one of the 

 faces strengthen the assumption of its association 

 with ice-action. The rock is of the usual type — a 

 very hard siliceous sandstone, white within and 

 stained externally by contact with ferruginous water. 



C. Carus-Wii-son. 

 October 10. 



NO. 2818, VOL. I 12] 



Dr. Kamtnerer's Ciena Experiments. 



In Nature of May 12, p. 639, Dr. Kammerer 

 wrote : " Not content with any of the previous experi- 

 ments [made by himself on the inheritance of acquired 

 characters], I carried out, before 1914, what may 

 really be an cxperimentum crucis," and Dr. Kammerer 

 states that when the siphons of Ciona intestinalis are 

 cut off they regenerate longer than they previously 

 were, repeated amputations giving very long tubes, 

 and that the offspring of these individuals have 

 siphons longer than usual. 



I repeated these amputation experiments between 

 June and September fast at the Roscoff Biological 

 Station. The oral siphon was removed from 102 

 Ciona intestinalis which were growing attached to 

 the walls of the tanks. The animals varied in length 

 from o-t) to 4'8 cm. .\s controls, 235 unopcrated 

 individuals were kept under observation. In none 

 of the operated animals was there any further growth 

 of the siphons after the original length had been 

 re-attained. 



One operation was performed on 59 individuals, 

 two on 35, and three on 8. The time necessary for 



S 2 



