21 On Unilateral Conductivity in Tourmaline Crystals. 



inductive capacity of a cold crystal in opposite directions along 

 its axis. 



So far, then, as experiment Las at present gone, the results 

 for the case of electricity have been altogether negative. For 

 the case of heat, distinct results have been obtained ; but it 

 is to be remembered that a rising temperature has always been 

 used, and no experiments on the conductivity for heat have 

 been at present made with the temperature constant : it is 

 probable that, when they are, negative results will be also ob- 

 tained corresponding with the negative res alts for electricity. 

 We have roughly repeated the heat experiments with & falling 

 temperature, and have obtained the ratio of conductivities in- 

 verted. 



Note by the second-named Author. 



It thus appears that our original hypothesis with regard to 

 the cause of the internal polarity of the particles of pyroelec- 

 tric crystals, at any rate in the form in which I put it forward 

 in section 26 of the paper " On a Mechanical Illustration of 

 Thermoelectric Phenomena" (Phil. Mag. Dec. Suppl. 1876), 

 has not been confirmed by experiment. 



But instead of this, the important result has been obtained 

 by Professor Thompson that, in a pyroelectric crystal whose 

 temperature is rising, Heat flows more easily with the Elec- 

 tricity (i. e. from the antilogous towards the analogous pole) 

 than it does against the electricity. This " convection of Heat 

 by Electricity " has an apparent analogy with the effect pre- 

 dicted and verified by Sir William Thomson in unequally 

 heated metals (Bakerian Lecture, 1856), and which might 

 equally well be called the convection of Electricity by Heat ; 

 and it must have an interesting bearing on the theory of Prof. 

 Kohlrauseh concerning Thermo-electricity and Heat-conduction, 

 set forth in Pogg. Ann. vol. clvi. p. 601. 



As to the original hypothesis, I am unable to give it quite 

 up even now. For though I have no faith in unilateral con- 

 ductivity in isotropic conductors like metals (upsetting as such 

 a thing would be to Ohm's law, which has been accurately 

 verified), yet in hemihedral crystals it did seem very possible 

 that greater resistance should be offered to the motion of elec- 

 tricity in one direction through them than in the other, just 

 as an ear of rye-grass is rougher one way than the other — that 

 there should in fact be something analogous to barbs, or valves, 

 so that when a non-directional disturbance (like a uniform rise 

 of temperature) was imparted to the crystal the electricity 

 should be urged from u to /3 more strongly than in the reverse 

 sense. Professor Maskelvne once told me that some crvstals 



