April 15, 1909] 



NATURE 



187 



chamber was closed in each case for half an hour before 

 observations of velocity were made. 



From the results just given it is clear that the figures 

 obtained by observation lor the rate of fall of the spores 

 are of the same order of magnitude as those demanded 

 by Stokes's law. However, the law is not confinned in 

 detail, for, as an average of the three experiments, it 

 was found that the actual velocity of fall of the spores 

 was 4O per cent, greater than the calculated. I have not 

 been able to find any satisfactory explanation for the dis- 

 crepancy between observation and theory. 



My method for testing Stokes's law appears to have 

 various advantages over that used by Zeleny and 

 McKeehan, for the following reasons : — Amanitopsis spores 

 have smooth walls, and are practically truly spherical, 

 whereas lycopodium spores have sculptured walls, and are 

 four-sided. Amanitopsis spores have a diameter only one- 

 third as great as lycopodium spores. In the tube method 

 convection currents cannot be eliminated, and it must 

 surely bo somewhat difficult to decide the exact centre of 

 the spore clouds. By my method of using a very small 

 chamber the difficulty of convection currents was reduced 

 so as to be negligible, and the velocities of the individual 

 spores could be measured with considerable accuracy. 

 .Amanitopsis spores are liberated spontaneously by the 

 fungus, whereas lycopodium powder requires to be set in 

 motion by artificial means. 



In conclusion, I wish to thank Prof. J. H. Poynting for 

 permitting me to carry out the experiments here recorded 

 in the physics department of the University of Birming- 

 ham, and also Dr. Guy Barlow for valuable criticism. 



h.. H. Reginald Buller. 



The Botanical Department, University of Manitoba, 

 Winnipeg, March 25. 



lonisation by Rontgen Rays, 



The relative ionisations produced in different gases by 

 beams of X-rays have been found by many investigators 

 to depend so markedly on the penetrating power of the 

 X-rays used that no regularity in behaviour has been dis- 

 covered (see Mr. Crowther's paper " On the Passage of 

 Rontgen Rays through Gases and Vapours," Roy. Soc. 

 Proc, January 14). 



Recent experiments which I have made upon homo- 

 geneous beams have, however, shown the connection 

 between ionisation, secondary radiation, and absorption in 

 a most striking way. As in the case of absorption pheno- 

 mena (see letter to Nature, March 5, Barkla and Sadler), 

 a knowledge of the secondary radiation characteristic of 

 an element is essential and sufficient to explain many of 

 the phenomena of ionisation. 



In order to test if such a connection existed, the first 

 substance experimented upon was ethyl bromide — a sub- 

 stance which has been investigated in some detail by Mr. 

 Crowther. 



By using homogeneous beams of X-rays, I found that 

 all radiations experimented upon which are not more pene- 

 trating than the secondary radiation characteristic . of 

 bromine (coefficient of absorption in .'\I = about 50) produce 

 ionisations which are proportional, or at least approxiinately 

 proportional, to the ionisation produced .by the same beams 

 in air. 



When the radiation passed through the vapour was made 

 more penetrating than the radiation characteristic of 

 bromine, the ionisation rapidly increased — that is to say, 

 the ratio of the ionisation in ethyl bromide to that in air 



NO. 2059, VOL. 80] 



rapidly rose to several times its original normal value. It 

 was found to be essential to the production of what may 

 be called the abnormal ionisation simply that the primary 

 radiation be more penetrating than the secondary radiation 

 which bromine emits. This result must be connected with 

 the results of experiments on absorption and secondary 

 radiation. 



Thus, when an X-radiation incident on a substance R 

 is softer than the secondary radiation characteristic of R, 

 it is absorbed according to a simple law, the absorption 

 being approximately proportional to the absorption in any 

 other substance in which a characteristic radiation is not 

 excited ; it produces no appreciable quantity of this 

 secondary radiation, and it produces what may be called 

 a normal ionisation in R. When the incident radiation 

 becomes more penetrating than the secondary radiation 

 characteristic of R, it is absorbed by an amount greater 

 than given by the law stated ; it begins to excite the 

 secondary radiation in R, and it produces an increased 

 ionisation in R. The absorption and ionisation increase 

 to several times their previous value, while the intensity 

 of secondary radiation becomes very great. 



.As the penetrating power of the incident radiation is 

 increased still further, the absorption by R diminishes, and 

 the secondary radiation e.xcited in R . diminishes at the 

 same rate as the ionisation produced by the incident radia- 

 tion in a thin film of air. 



(It should be pointed out that the great increase in 

 ionisation is not due to the secondary radiation.) 



In a similar manner, from a knowledge of the secondary 

 X-rays emitted by iodine, the variable behaviour of methyl 

 iodide may be explained. The effects of the lighter 

 elements are comparatively small in all the three pheno- 

 mena of absorption, secondary radiation, and ionisation. 



Very many of the apparently complex results, obtained 

 by experiments on the transmission of heterogeneous beams 

 through compound substances, may be explained in terms 

 of a few simple laws which have been obtained by the 

 more fundamental experiments on elementary substances 

 with the use of homogeneous beams. 



* Charles G. Barkla. 



University of Liverpool, April 7. 



A Simple Fabry and Perot Interferometer. 



DcRiNG a course of experiments with interferometers 

 it was found that a very simple and inexpensive Fabry 

 and Perot instrument could be constructed of plate glass 

 which gives results almost as good as the costly interfero- 

 meter. The construction of this apparatus for demonstra- 

 tion purposes will well repay the teacher and student. 

 The sharp-coloured interference rings obtained by using 

 luminous gases in vacuum tubes as sources are extremely 

 beautiful. The D lines from a sodium burner are easily 

 separable. If the interference pattern, using a copper or 

 iron arc, is focussed on a wide slit of a single-prism 

 spectrometer, a section of the interference rings is seen 

 in, the various spectrum lines, illustrating the method of 

 Fabry and Buisson, and Eversheim, for the determination 

 of the new standard table of wave-lengths. The Zeeman 

 effect can also be easily shown with this apparatus. 



Take two pieces of plate glass about an inch square (I 

 have used the so-called German plate) and silver ' them 



until one surface of each plate cuts down the intensity 

 of the transmitted light to about a quarter of the incident 

 light. Separate these silvered surfaces by two strips of 

 cardboard. A useful ' thickness to begin with is about 

 045 mm., as this , will clearly separate the D lines. 

 Mount these plates over a half-inch hole in a metal plate 

 by means of three pressure screws, t\yo, of whi^rh are 

 shown in the above diagram, being a section through 



I For silvering soluttc 



; the appendix to Baly's " Spectroscopy. ' 



