Vol. 4] Thehn.— Thermal Conductivities of Certain Schists. 213 



A word as to the kind of point filed on the wire will not 

 be amiss. At first it was the tip of a. long cone, but no heat 

 can flow across a point, and the tip of the cone had to come off. 

 The question was how much of it, how large a section would 

 give the best results. Without going into the details of differ- 

 ential heat equations, there are in general two limits to be dis- 



Fig. 3. 



cussed, — a very large section or a very small section. In the 

 case of the latter (fig. 4), the only objection is that the heat 

 which flows across can be dissipated from the surface of an 

 ellipse which is very small indeed. In the case of the former 



Fig. 4. Fig. 5. 



(fig. 3), both x and y would be greater, but a good contact of 

 the wire on the rock could hardly be expected. Figure 5 indi- 

 cates roughly what might be obtained in the way of a contact, 

 where the unshaded portion would show an air film between 

 the copper wire and the rock. Then, too, a large section that 

 is as nearly circular as it can readily be fashioned, would give 

 appreciably different values for the diameter in different direc- 

 tions, and d would not be a true constant. Another element 

 which might be worth considering is that, in measuring many 



