e334 . Scientific Intelligence, 



where is the angle made by jP§ with the axis of z, and r denotes 

 the distance from P to Q. It is shown that, under these conditions 



— = — = =: ■\- cc . Hence, v V has no mean- 



c)x dy"^ 9z^ 



ing, and Poisson's formula fails at P. 



Case II. When p = cos/log— j it is proved that the three 



second derivatives .of the potential are all non-existent, so that 

 Poisson's equation again breaks down. 



Case III. When p = cos — all of the second derivatives of V 



r 



are equal to zero, consequently Poisson's equation is not satisfied 



unless p^ = 0. *'It should be noted that, in this case, p^ may be 



assigned any value without affecting the value of v^ F." 



The last section of the paper is devoted to proving, for the first 

 time, that Petrini's generalization of Poisson's equation is of 

 limited scope, and fails for case II. — Phil. Mag.^ xxxiv, p. 138, 

 August, 1917. H. s. u. 



8. The Composition of J^-Pays froin Certain Metals. — The 

 composition of the X-radiations emitted by anticathodes made of 

 aluminium, copper, iron, nickel, and platinum has been investi- 

 gated by G. W. Kaye. The qualitj^ of the JT-rays was tested by 

 placing absorption screens between the aluminium window of the 

 bulb and the ionization chamber. A Wilson tilted electroscope 

 Avas employed to measure the ionization currents. The experi- 

 mental data were plotted with thickness of absorbing screen d as 

 abscissa and with the corresponding value of log,o(/d/^o) as ordi- 

 nate. Since 1^-=. I^ e~^, it follows that the absorption-coefficient 

 A., at any point on the graph, is obtained by multiplying the slope 

 of the tangent to the curve at the given point by 2*3 (the recipro- 

 cal of the modulus). For heterogenous X-rays A diminishes as d 

 increases, but for homogeneous rays the graph is a straight line. 

 The plan followed in the experiments was gradually to cut down 

 the X-rays until the log-absorption curve indicated that the resid- 

 ual rays were homogeneous or nearly so. This " end-radiation " 

 was subtracted graphically from the total radiation by prolonging 

 backwards the final straight segment of the absorption curve. A 

 second log-absorption curve was then plotted, employing as ordi- 

 nate the logarithm of intensity-difference derived as stated. The 

 resulting graph proved in practice to be either a straight line 

 throughout its length, or a curve merging into a right line as the 

 thickness of the screen increased. In the latter event, the process 

 of subtraction was repeated. 



The following useful facts were brought out by the investiga- 

 tion. The X-rays from a bulb excited by low voltages (10,000 

 to 50,000 volts) are rich in the characteristic radiation of the auti- 



