ON ARTIFICIAL REFRIGERATION. 949 



■with a mmimnm amoinit of the rofti^CTaiifc circulating in tho mn^Hiino. 

 Through a series of long tubes, fixed at their ends in tube-plates in tho 

 ordinary manner, I run a corresponding series of tubes, but of a smaller 

 diameter and slightly longer; and tlie ends of the latter are likewise 

 fixed in tube-plates, and the spaces between these tube-plates are made 

 to form closed chambers. The interior of these chambers, at the differ- 

 ent ends of the pipes, are connected wnth one another by the thin an- 

 nnlar sj^aces wdiich intervene between the inner and outer tubes. The 

 refrigerant is caused to flow from one chamber to the other through these 

 annular spaces between the pipes, while the liquid to be cooled passes 

 through the small tubes and over the outer ones. In the condenser the 

 ■warm gas, iTuder compression, occupies the annular sjiace, and a current 

 of condensing water flows fi-eely through the bore of the small tubes and 

 outside the larger ones. 



The ends of the outer tubes are expanded into the inner tube-plates, 

 and the smaller inner tubes are secured at both ends in outer tube^ 

 plates by stufiQng-boxes, so that the inner tubes can be withdrawn and 

 the whole apparatus cleaned. This is an advantage possessed by no 

 other machine. 



The smaller chamber, occupied by a vertical coil leading from the 

 condenser to the refrigerator, represents my supplementary condenser, 

 in which the temperature is regulated so as to determine at will the 

 pressure at which the machine is to be operated. This manifest improve- 

 ment, in any form of apparatus in which gases are liquefied by compres- 

 sion, was suggested to me by the varying conditions observect in work- 

 ing freezing-machines on board ships sailing from temperate to tropical 

 climates. The warm condensing water available at the equator led to 

 such pressures as to interfere most materially vrith the economical work- 

 ing of the machines. By using a flow of the cooled, uncongealable liquid 

 returning from the ice-boxes or cooling chambers, it is easy to obtain, at 

 some cost of i^ower, a regular temperature of 32° Fahr. or under, so as 

 to prevent undue and dangerous pressures. 



It will be noticed that so far, all freezing machines in which a liquid 

 is volatilized, circulated, and recondensed by a pump, consist of four 

 fundamental parts, ^dz, engine, pum^i, condenser, and refrigerator, and 

 the imj)rovements from time to time on Jacob Perkins's machine are im- 

 provements in detail of construction, the most imj)ortant of which have 

 rendered the machines more comjiact and efficient. 



Every machine so far has worked with most economy, the colder the 

 condensing water and the more perfect the arrangements for complete 

 liquefaction, on the one side, and ready evaporation, by supply of heat, 

 on the refrigerator side. AU, except the Carre machine, w hich presents 

 serious drawbacks as compared with any apparatus in which anhydrous 

 ammonia is alternately liquefied and volatilized, have been wasteful iii 

 proportion to the inefficiency and inevitable loss of heat by the steam 

 engine, and these considerations led me to design my 



