Jantiakt 15, 1915] 



SCIENCE 



79 



indicated by Bohr's theory, the agreement 

 was found to be a remarkably close one. 



With elements of higher atomic weight 

 Moseley obtained spectra whose lines indi- 

 cated the Barkla "L type" of radiation. 

 The atomic numbers calculated from the 

 positions of the strongest lines of these 

 "L" spectra ranged from 40 for zirconium 

 to 79 for gold. These experiments then 

 give strong support to the hypothesis of 

 van der Broek that the total charge of the 

 electrons of an atom indicates its position 

 in the periodic system. Known elements 

 were found to correspond with all the num- 

 bers from 13 to 79 except three, indicating 

 that three elements probably remain to be 

 discovered. The wave-lengths of the char- 

 acteristic X-rays from the metal is of the 

 order of 1/1000 that of visible light (i. e., 

 about 40 waves in .000001 inch). 



During the past few months Rutherford 

 and Andrade^* have extended these meth- 

 ods of crystal reflection to the study of 

 radiation from Ra-B Ra-C. The y-ray 

 spectrum of Ra-B was found to be of the 

 same general type as that of the X-ray 

 spectrum from various heavy metals when 

 bombarded by cathode rays. The result 

 for soft y-rays from Ra-B shows that its 

 radiation belongs to the "L series" for 

 heavy metals. Moseley 's formula applied 

 to the measurement of the lines of the y-ray 

 spectrum gave N = 82, which is the atomic 

 number of lead. The atomic weight of 

 Ra-B is, however, 214, while that of lead 

 is 207. This difference is nevertheless fully 

 explained by a new generalization of Soddy 

 and Fajans which we will presently notice. 

 The experiments described in the second 

 paper were made with much more penetra- 

 ting y-radiation from both Ra-B and Ra-C. 

 This penetrating y-radiation from Ra-B was 

 found to correspond to the K series for the 

 same metal, lead. The still more penetrating 

 radiation from Ra-C has a line spectrum 



of still higher frequency than the K type, 

 for which the name "H" series is sug- 

 gested. These rays are especially interest- 

 ing because they have by far the shortest 

 wave-lengths yet known, only about 1/8 of 

 the wave-length of the shortest X-ray waves 

 measured by Moseley or about 1/80,000 of 

 the wave-length of sodium light. Ruther- 

 ford in his comments on these waves very 

 justly remarks, "It is surprising that the 

 architecture of the crystals is sufficiently 

 definite to resolve such short waves." 



During 1913 some remarkable work on 

 the relations of radioactive substances to 

 each other has given support to the nucleus 

 atom from an unexpected quarter. Fleck,^^ 

 Russell,^" Von Hevesey,^' Fajans^* and 

 Soddy^^ have all had a share in this work. 

 They have found that when a radioactive 

 substance ejects an a-particle a substance 

 of different chemical properties and differ- 

 ent valency results. The new substance 

 lies two columns to the left in the periodic 

 table, has an atomic number two less and 

 an atomic weight about four less than the 

 parent substance. If however the radioac- 

 tive substance ejects a ;S-particle or elec- 

 tron, the new substance is one column to 

 the right in the periodic table, increases 

 one in atomic number, and does not change 

 in atomic weight. Plainly then two or 

 more elements may occupy the same posi- 

 tion in the periodic table, for if an element 

 loses in succession — in any order — two /S- 

 particles and one a-particle, its atomic 

 number will be again the same as it was at 

 first. Thus Ra-D has the atomic number 

 82 ; it loses a /3-particle and becomes Ra-B 

 with atomic number 83,; this loses another 

 /3-particle and becomes Ra-F with atomic 

 number 84; this finally loses an a-particle 

 and becomes lead, with the original atomic 

 number 82. The series Url, UrXl, 

 Ur X 2 and Ur2 is of the same kind, except 

 that the particles are ejected in the reverse 



