288 Mr. Norman Campbell on Delta Rays. 



before it can be definitely accepted a most important source 

 o£ uncertainty which has been left unnoticed hitherto must 

 be discussed ; the discussion will be postponed till a few 

 farther facts have been recorded. 



12. The third row of Table I, should represent the charge 

 carried by the alpha rays themselves, and should, therefore, 

 measure the number of alpha rays acting to produce delta 

 rays. The numbers actually given require some correction, 

 for, even in the absence of the polonium disk, the plate A 

 tended to charge positively. Many blank experiments were 

 made to determine this residual current, but the results 

 obtained were not very consistent : the quantity which re- 

 quires to be subtracted from the values in row 3 in order to 

 obtain the true charge of the alpha rays is about 0'04, but 

 the uncertainty is such that the resulting values cannot be 

 relied upon in column 1 to 10 per cent., in column 9 to 

 100 per cent. This uncertainty, since it is the same for the 

 whole curve, does not affect any of the other conclusions 

 drawn. But the comparison of the total amount of alpha 

 rays obtained in this way with the total amount of delta rays 

 obtained by adding the corresponding + and — columns in 

 row 4 lead to an interesting conclusion, the main features of 

 which are not dubitable. In the last row of Table I. the 

 ratio of the charge on the delta rays liberated to that on 

 the exciting alpha rays is given : it must be remembered 

 that the alpha rays carry a charge double that of the delta 

 rays. It is clear that one alpha ray can liberate more than 

 one delta ray, and that the number it can so liberate is of 

 the order of 10. Apparently, then, the delta rays must come 

 from a layer about 10 atoms thick. Further, it will be noted 

 that the slower rays in general liberate more delta rays than 

 the faster. This conclusion is in accordance with that 

 derived from the study of ionization ; indeed, it is the one 

 conclusion attained in these experiments which might cer- 

 tainly have been anticipated. 



Again, if the simple theory is correct, the difference 

 between the values of "Y = 0" and "V=0 (magnet)" should 

 represent the excess of the number of delta rays tending to 

 charge the plate A positively over those tending to charge it 

 negatively ; that is to say, it should be equal to the excess of 

 the value in the fourth row marked — over that marked 4- . 

 (In thinking out the signs it should be remembered that the 

 — column represents the properties of the rays which are 

 stopped by a negative potential on B and those, therefore, 

 which tend to charge A positively.) The relation is not 

 fulfilled, the difference between the numbers in row 7 4 beino- 



