REFRIGERATION AND REFRIGERATOR INSULATION ON BOARD SHIP. 151 
One cannot conclude from the suggestion that cork can be cemented directly 
on to the steel in the construction of refrigerator compartments on board ship that 
a similar construction is suitable for ice-tank work in stationary plants. In this 
work the surfaces of the tanks are, in general, not particularly true. There is also 
the difficulty of rivet heads which must have the cork fitted about them. In work 
of this size this fitting is apt to be done carelessly, with the result that air spaces will 
occur and the cork be forced out of place by freezing and melting of moisture. 
Difficulty has been encountered, in one or two installations which the writer 
has seen, with the use of cork that had not been sufficiently pressed previous to bak- 
ing. This cork was not as dense as it should have been and had not the mechanical 
strength it should have had. More or less air spaces were left in the cork, into 
which there was, of course, the tendency for moisture to penetrate, and where the 
temperature was very low this moisture froze and pried the sheets apart, completely 
disintegrating the insulation and requiring that the entire work be done over again. 
The idea seems to prevail that a refrigerator door should have beveled edges 
and that these beveled edges make a very tight door. How this idea originated it is 
beyond the writer’s power to divine. Asa matter of fact, such a door can be tight 
only if it happens to be very carefully fitted. As soon, however, as the moisture be- 
gins to have its inevitable effect, the door cannot by any chance be tight. Further- 
more, if the hinges yield at all, as they are certain to do with a heavy refrigerator 
door, it will never be tight, even though it does not warp at all. 
In the writer’s opinion there is only one practical way to make a tight refrig- 
erator door, that is, to have a plane surface come up against a plane surface with 
a flexible gasket between these two surfaces, as shown in Plate 76. It may 
also be possible to have a pair of such surfaces, or, rather, two pairs of such sur- 
faces with gaskets between them, making an air pocket between the two seals. If 
this arrangement is used the door can sag considerably without causing any serious 
binding or failure in tightness. 
This matter of a satisfactory refrigerator door cannot be emphasized too 
much. Our experience has been that refrigerators built by some of the best known 
shipbuilders have been fitted with extremely bad doors, thus seriously increasing the 
difficulty of properly refrigerating the boxes. 
REFRIGERATION NOTES. 
In estimating the capacity of a refrigerating machine required for taking care 
of a given refrigerator, it is not sufficient to calculate the heat loss through the in- 
sulation of the refrigerator. In fact, this loss in many cases is very much smaller 
in proportion to the total loss than the average engineer realizes. 
It is proportionately so small, in fact, in the case of ordinary household refrig- 
erators that almost all the builders of this class of refrigerator, after comparing 
the ice consumption in well and in poorly insulated boxes, use very low grade in- 
sulation in their stock boxes. Their custom, in general, is to reduce the size of the 
air passages leading to and from the ice compartment to a very small area and 
