CONTEMPORARY ADVANCES IN PHYSICS 169 



Halle) measured at short intervals over the length of a week; the 

 difference is, that the lowest shows the effect inside armor of lead ten 

 cm. thick, while in getting the data for the topmost the plates of lead 

 were removed from above the apparatus, though left in place at the 

 sides and below. The lower curve thus represents the action of hard 

 cosmic rays: one sees that it rises as the barometric pressure falls, 

 falls as the latter rises — the heavier the blanket of air between Halle and 

 the sky, the more these hard rays are reduced on their way to the 

 ground. The upper shows the joint action of the hard cosmic rays, of 

 others which are softer, and of soft rays from radioactive substances in 

 the air: one sees how violently it fluctuates, supposedly (in Hoffmann's 

 view) because the radioactive atoms wandering in the air change 

 greatly in number as the weather changes. Figure 10 shows values of 

 ionization obtained high up (3500 metres) in the Alps, on various 

 August and September days of 1927, plotted against sidereal time; the 

 zigzag lines connect observations made on two particular days. 

 Corlin is said to have inferred, from a statistical study of data obtained 

 in northern Sweden, that the ionization ascribed to cosmic rays 

 decreases gradually and slowly before a magnetic storm, leaps suddenly 

 to a high value at the onset of the storm, then decreases again. ^ 



There is a related question which has been much debated : does the 

 intensity of the cosmic rays vary as different celestial objects pass 

 overhead, the sun for instance, or the Milky Way? As the years go 

 on, the answer to this question becomes steadily more and more 

 strongly in the negative. The opposite opinion has been held largely 

 by German physicists ; but a year ago one of them (Hess) reduced the 

 proportion of the cosmic rays which he considers dependent on the 

 sun, to half of one per cent of the total amount. 



Certain perplexing data must be mentioned before we go on to the 

 work which is done with other instruments than the ionization- 

 chamber, for they impeach the reliability of this device. From many 

 experiments of the pupils of Swann, it appears definitely established 

 that if the quantity of air in an ionization-chamber is increased, the 

 number of ions appearing in that chamber in unit time increases in a 

 lesser ratio, and in fact approaches (and in some practical cases, even 

 attains) a limiting value (Fig. 11). At first this seems natural enough: 

 the limiting value should be attained when the air is so dense, that the 

 ionizing rays are entirely absorbed before they have completely crossed 

 the chamber. But, appreciable fractions of these ionizing rays are 

 able to penetrate many metres of water, many centimetres of lead : it 

 is pretty nearly a formal contradiction-in-terms, to assert that they 



8 See \V. M. H. Schulze, Nature 128, 837-838 (14 Nov., 1931). 



