478 DR. CHARLES G. BARKLA ON POLARISED RONTGEN RADIATION. 



of thinner pulses the action in copper might more nearly approximate to that taking 

 place in light substances during the passage of thicker pulses. The more easily 

 absorbed portion of the secondary radiation was therefore cut off from the secondary 

 electroscopes by plates of aluminium '04 centim. thick, and in many experiments 

 there appeared to be a slight variation as with lighter substances. This, however, 

 was within the limits of possible error of experiment. Also it must be remembered 

 that there was a radiation from air superposed on that from copper, and though it was 

 negligible in comparison when the radiation was unintercepted by any absorbing plates, 

 yet when the aluminium plates were used the radiation from copper was absorbed to 

 a much greater extent than that from air and the resultant effect from air became 

 appreciable ; hence the slight variation may be accounted for. 



Similar experiments were made with other heavy metals, but no evidence of 

 variation in intensity of secondary radiation was obtained when the bulb was turned 

 about the axis of the primary beam. 



Regarding the amount of variation of intensity of the secondary beam as the bulb 

 was turned, it will be seen that from the light substances it was 11 or 12 per cent, of 

 the total intensity. This is not a true measure of the amount of polarisation, for 

 beams of considerable cross section were studied, consequently secondary rays making 

 a considerable angle with the normal to the direction of propagation of the primary 

 rays were admitted into the electroscopes. 



Different experiments with a given bulb and given radiator in a fixed position 

 appeared to indicate polarisation to a slightly different extent as the bulb was worked. 

 It is possible that the amount of polarisation depends to a certain extent on the 

 method of discharge through the X-ray bulb, but the variation was not sufficient to 

 absolutely prove this. 



The variation between the amount of polarisation shown by using air, cardboard, 

 and aluminium as radiators was not outside the limits of experimental error. 



The conclusions from these experiments may be briefly stated thus : 



Partial polarisation exists in a beam of Ro'ntgen radiation proceeding from an X-ray 

 focus tube. 



The intensity of secondary radiation from air and light solids in a direction 

 perpendicular to that of propagation of the primary radiation depends on the position 

 of the plane of polarisation of that primary radiation. 



The intensity of secondary radiation from the heavier metals is independent of the 

 position of the plane of polarisation of the primary radiation. 



[Variation in the penetrating power of the primary radiation has not been observed 

 to affect this result.] 



[NOTE, April 15, 1905. Since writing the above paper I have obtained a primary 

 X-radiation which gives rise to secondary radiation differing in intensity in the two 

 principal directions by about 20 per cent. ; consequently the possible error has been 



