12 Mr. C. Chree on the 



Weber next considers the liquid layer separately ; and his 

 work, though of a somewhat approximate nature, is fairly 

 satisfactory. His conclusions are practically in accord with 

 those deduced by Lorberg from a very able and complete 

 treatment. Lorberg considers the case of n horizontal disks 

 of the same radius, but of different materials, one above 

 another, all originally at one temperature. The lowest is 

 suddenly put on an ice-plate, and the whole surrounded by 

 an atmosphere at 0°. The problem is very complicated, 

 especially as Lorberg is driven to attach different meanings 

 to the same letter in different parts of the work, and has to 

 use an elaborate system of suffixes. He applies the results 

 to three disks — copper, water, copper — in order, but compli- 

 cates matters still further by inverting his previous notation. 



The conclusion come to both by Weber and Lorberg is 

 that, when cooling has lasted a comparatively short time, the 

 temperature of the hot junction is given with sufficient 

 accuracy by a single term, 



According to the former, 



while the latter gets 



pc* 



pG x SX pC 



w T here R is the radius of either plate. Neglecting smaller 

 terms, Lorberg's equation for q is, in the present notation, 



PiCia a i A * 90 h(2 , 1\ , A 2 h 2 



^^Aitan^A = l ^ — r k+ aT +~2 7r' 



pG q p x Ci k \R AiJ q* k R 



while Weber's differs only in leaving out the term in h. 



In Lorberg's equation (25), from which the above equation 

 is taken, there are several terms of a lower order. These are 

 deduced from his general equation (22), with which I entirely 

 agree. I further coincide with the approximate equation 

 given above, but after carefully working through all the 

 algebra, which Lorberg leaves the reader to supply, am 

 unable to agree entirely with these secondary terms. The 

 discrepancy, however, would affect the value of q only to a 

 very small extent. Lorberg finally gives 



^ = C231)- 2 { ' 



•2539--1472 



pc 



= 4*758 — term in /i ; 



