LOW TEMPERATURES AND REFRIGERATION—MARCHIS. 219 
attained an equilibrium the temperature is taken at different points 
along the bar, and by the formule of Fourrier the coefficient of con- 
ductibility can be calculated, if that of emission is known. This 
jatter may be that of a coating or a very thin layer of metal, with 
which the bar has been previously provided. 
A variation of the Forbes method is that of the sectional bar of 
Lodge. This bar is composed of three sections; the first and the 
third are of a metal the conductibility of which is known. The 
second is made up of the material to be tested. The end of the first 
section is heated and the progression of temperature when equilib- 
rium is established is measured. The general formula for uniform 
movement of heat in an elongated bar makes possible the calculation 
of the conductibility of the material composing the second section. 
It is this method that M. Desvignes has used in determining the 
conductibility of several insulating substances. He has worked out 
the technique so that it can be easily employed in the refrigerating 
industry. 
Some of the results obtained by this method are as follows: 
Coefficient of conductibility. 
Calories for 
meter-degree-hour. 
OT cae ee a rh re es = ee a ee 0.05 to 0. 014 
(GinsianGre “ood a a ee a ee eee . 068 
Cork withwecaseinkbind ers a a eee ee ATL Pon Bee MEE RY . 069 
Corkswith odorless: pitch) bind erses es eee ee ae . OST 
(Cloke Gyalitin Storeliionen bias loniavemye . O67 
Cork mixed with infusorial earth and calcined_________-~----___ . O89 
As M. Desvignes has remarked, it would perhaps be imprudent to 
take these figures as a basis for the calculation of loss in a cold-stor- 
age plant. The specimens which were used in the tests were picked 
and were perfectly dry. The given coefficients should be increased 
not less than 20 per cent. In the application to certain materials 
some account should be taken to the method of joining. Thus in a 
partition of cork bricks pointed by cement mortar, where the joints 
represent in very careful work about 15 per cent of the total volume, 
the coefficient of conductibility of the brick itself, referred to the 
total surface of the partition, should be almost doubled. These con- 
siderations will show what a difficult question even the approximate 
calculation of the heating effect due to the walls of the refrigerator 
may be. The second section has withheld opinion on this question, 
therefore, and taken the following resolutions: 
(1) That study and research shall be undertaken in the technical 
schools and laboratories to determine, either by known apparatus or 
that which shall be subsequently devised, the specific constants of the 
different insulators which are practically utilizable in the refriger- 
ating industry. 
