AMERICAN INSTITUTE. 595 



The methods which an observation of these principles would 

 suggest, have been often adopted, but more frequent from conven- 

 ience than an intelligent regard to tlie reasons of their efficiency. 

 Wooden walls are nearly always made double, but more reliance 

 being placed upon the wood and the plaster than upon the air 

 between them, care has not been taken to make the enclosed 

 spaces air tight, especially at the top and bottom. Double win- 

 dows are generally supposed to derive their value from the addi- 

 tional thickness of glass solely. The common custom of placing 

 a temporary sash upon the outside, has the objection that the 

 windows cannot be opened. A better plan is to make the sash 

 thicker than is usual, placing permanently one thickness of glass 

 near the outer edge, and inserting a thin sash of hard wood, con- 

 taining another plate of glass, upon the inside, fastening it with 

 screws, so that it may be removed for cleansing the interior sur- 

 faces. If both plates are permanently secured with putty, the 

 glass will in time become dimmed, either from oxidation or the 

 condensation of the impurities of the confined air. 



This arrangement would be specially valuable in conservato- 

 ries, graperies, &c., where it is occasionally desirable to open the 

 sashes. In railroad cars, too, it would prevent the annoyance of 

 frost upon the windows, which so often obstructs the view in 

 the winter season. 



Another class of structure, from which all external heat is to 

 be excluded, has given occasion for the display of much inge- 

 nuity, and to this the principles laid down are peculiarly appli- 

 cable. Formerly ice houses were invariably built under ground. 

 This reduced their temperature to that of the earth around them, 

 but as that was found to be a better conductor than even the 

 unconfined air, it was exceedingly difficult to maintain at any 

 lower point. Of late years tlie favorite plan has been to erect 

 them above ground, and fill them in with straw, tan, saw-dust, 

 or other substances, whose chief value consists in confining the 

 air held within the interstices. If, however, the filling itself 

 becomes damp by any means the walls are made to acquire a con- 

 ducting power equal to that of the water they contain, and a loss 

 of ice must always be calculated upon from this cause. For 

 domestic uses, an ice-house may readily be made below the 

 earth much more efficient than any other in use, by enclosing a 

 pit with cemented walls, covered with an air-tight roof. In this, 

 and a few inclies from it at all points, a tight wooden box is to 

 be erected, leaving an air space all around it, as well as above 



