296 MAGNETISM 



where T is the absolute temperature. If we .i^mne that this holds 

 when the oxygen is liquefied at - 182 C. or + 91 A., and if 

 take the density of liquid oxygen as 1*14, thi-n 



y i 



The change from gas to liquid, then, makes no great change 

 in the value of \. 



Pascal's experiments. Pascal* studied the magnetic quali- 

 ties of a large number of solutions, of some liquids, and 

 liquefied gases by a method due to Quincke. In principle tin- 

 liquid to be experimented on was contained in a U-tubc with 

 limb narrow {(j mm. diameter), the other wide (Gem. di.t 

 narrow limb was between the poles of an electro-magi x 1>\ whit h 

 a horizontal field could be establishes! ot 'any intensity up to i 

 thousand gausses. The surface of the liquid in this limb was at 

 the centre of the field and it s 1. \ t 1 \\ ax <!,., ( \ , ,| \\ ith a microsi 



The field at the surface of the wide limb was in cnmp.t: 

 negligible. Hence when field II was put on. if the difl. 

 level in the limbs was //, if K is the Misn-|,t ibiht \ o; the liquid 

 /c of whatever is above it, the toice per unit area was as in Wills' 

 experiment : 



Hut this balances the column of height //. It' the dcnsjt\ of the 

 li(iuid is n, 



The actual change ob i the microscope is tliat of the 1< 



in the smaller limb, say f, If 8 be the cross-s<-( tion .: r row 



limb and S that of tin w ider, it is easy to see that h = (S-f *>5/S, so 



that 



To avoid measurement of H, the experiment was repeated on 

 water with the >amc value of II. 1 ) not i u ptihihtx 



by * its density by f > l% and the change of level in the nai 

 limb by x , 



By division 



* -I"". '/' i'1'ii.i. -M S ser. XVI. (1909). p. 3.';'. nnl XIX. (1910>, p. 6. 



