800 



SCIENCE 



[N. S. Vol. XL. No. 1040 



the general heterogeneous radiation two in- 

 tense almost homogeneous sets of waves. For 

 instance, rhodium gives two pencils of wave- 

 lengths, approximately equal to 0.61 X 10"* cm. 

 and 0.54 X 10"* c™- respectively. More exactly 

 the former of these is a close doublet having 

 wave-lengths 0.619X10"* and 0.614X10"*- 

 The wave-lengths of palladium are nearly 

 0.58 X 10"* and 0.51 X 10"* ; nickel 1.66 X 10"* 

 and 1.50 X 10"*- Lately Moseley has made a 

 comparative study of the spectra of the great 

 majority of the known elements, and has 

 shown that the two-line spectrum is character- 

 istic of all the substances whose atomic weights 

 range from that of aluminium, 27, to that of 

 silver, 108. These X-rays constitute, there is 

 no doubt whatever, the characteristic rays 

 which Barkla long ago showed to be emitted 

 by this series of substances. 



Now comes a very interesting point. When 

 Moseley sets the increasing atomic weights 

 against the corresponding decreasing wave- 

 lengths, the changes do not run exactly paral- 

 lel with each other. But if the wave-lengths 

 are compared with a series of natural num- 

 bers everything runs smoothly. In fact it is 

 obvious that the steady decrease in the wave- 

 length as we pass from atom to atom of the 

 series in the periodic table implies that some 

 fundamental element of atomic structure is 

 altering by equal steps. There is excellent 

 reason to believe that the change consists in 

 successive additions of the unit electric charge 

 to the nucleus of the atom. We are led to think 

 of the magnitude of the nucleus of any ele- 

 ment as being simply proportional to the num- 

 ber indicating the place of the element in the 

 periodic table, hydrogen having a nuclear 

 change of one unit, helium two, and so on. 

 The atomic weights of the successive elements 

 do not increase in an orderly way ; they raount 

 by steps of about two, but not very regularly, 

 and sometimes they seem absolutely to get into 

 the wrong order. For example, nickel has an 

 atomic weight of 58.7, whereas certain chem- 

 ical properties and still more its behavior in 

 experiments on radio-activity indicate that it 

 should lie between cobalt (59) and copper 

 (63.6). But the wave-lengths, which are now 



our means of comparison, diminish with abso- 

 lute steadiness in the order cobalt, nickel, 

 copper. Plainly, the atomic number is a more 

 fundamental index of quality than the atomic 

 weight. 



It is very interesting to find, in the series 

 arranged in this way, four, and only four, gaps 

 which remain to be filled by elements yet un- 

 discovered. 



Let us now glance at another and most im- 

 portant side of the recent work, the deter- 

 mination of crystalline structure. We have 

 already referred to the case of the rocksalt 

 series, but we may look at it a little more 

 closely in order to show the procedure of 

 crystal analysis. 



The reflection of a pencil of homogeneous 

 rays by a set of crystalline planes occurs, as 

 already said, at a series of angles regularly 

 increasing; giving as we say, spectra of the 

 first, second, third orders, and so on. When 

 the planes are all exactly alike and equally 

 spaced the intensities of the spectra decrease 

 rapidly as we proceed to higher orders, accord- 

 ing to a law not yet fully explained. This is, 

 for example, the case with the three most im- 

 portant sets of planes of sylvine, those per- 

 pendicular to the cube edge, the face diagonal, 

 and the cube diagonal respectively. An ex- 

 amination of the arrangement of the atoms in 

 the simple cubical array of sylvine shows that 

 for all these sets the planes are evenly spaced 

 and similar to each other. It is to be remem- 

 bered that the potassium atom and the chlorine 

 atom are so nearly equal in weight that they 

 may be considered effectively equal. In the 

 case of rocksalt the same may be said of the 

 first two sets of planes, but not of the third. 

 The planes perpendicular to the cube diagonal 

 are all equally spaced, but they are not all 

 of equal effect. They contain alternately, 

 chlorine atoms (atomic weight 35.5) only and 

 sodium atoms (atomic weight 23) only. The 

 effect of this irregularity on the intensities of 

 the spectra of different orders is to enhance 

 the second, fourth, and so on in comparison 

 with the first, third and fifth. The analogous 

 effect in the case of light is given by a grating 

 in which the lines are alternately light and 



