car. These slats arc not suitable for loading boxes. The fourth difference 

 consists in the form of partition between the ice bunker and the car proper. 

 In both cases there are spaces above and below the partition providing for 

 the flow of warm and cold air. But in the Hanrahan the partition, with 

 the exception of these spaces, is solid, and thus gives definite direction 

 to the air currents. In the Bohn system the partition is formed of gal- 

 vanized iron slats, placed like the slats in a window shutter. Seel ions of 

 the cars are show^n in Fig. 1, and illustrate these differences. 



It was expected that a transit of six or eight days with plums and 

 Crawford peaches would indicate any difference in efficiency between these 

 cars. It is claimed that some refrigerator cars fail to maintain a uniform 

 lernpernUire throughout the car — the top of the car and the parts most 

 remote from the ice, being, it is claimed, several degrees higher in tempera- 

 ture than the bottom near the ice- This difference of temperature would of 

 course be most pronounced in hot weather, and in cool weather would be 

 small. It so happened that the weather during this shipment was quite 

 cool. Whether or not this fact accounts for the uniformity in results can- 

 not be asserted, but at all events there was no noticeable diffet-ence in the 

 condition of the fruit upon arrival at Winnipeg. As a test of efficiency 

 in the cars, therefore, the experiment is inconclusive, and must be repeated 

 in warmer weather or over greater distances, before any decisive report 

 can be given out upon this matter. 



Fig. 2 —Section of car showing interior construction for air circulation with bunkers at end. 

 Trie solid pHrtition next to the ice and the norizontal flue overhead will, it is believed, improve the 

 efficiency of refrgerator cars. 



To maintain uniformity of temperature throughout a refrigerator it 

 has been already demonstrated beyond question that a regular circulation 

 of air through the whole system, including refrigerator and ice-house," is 

 necessary. To secure circulation the construction of the refrigerator must 

 be such as to give direction to the air. In principle, and in practice where 

 the test is sufficiently severe, the Bohn system may be improved by adopt- 

 ing the essential parts of the Hanrahan system where this relates to air 

 c'rculai^on; that is, a solid instead of a slatted partition, and a flue over- 

 head, (FV. 2). 



