POLENSKE PROCESS. 



247 



author's investigations lead him to conclude that a very important 

 factor, the temperature of the air around the flask and the rate 

 of conduction through the walls of the flask and bulb tube, is 

 left to chance, and he, therefore, proposes supporting the flask 

 on a piece of asbestos cardboard in which a hole 5 cm. in diameter 



Fig. 34. Steam Jacket. 



is cut, and surrounding the flask with a vessel in which water ts 

 boiled (see Fig. 34). Using this apparatus, it is found that with 

 butters the length of time of distillation can be varied to a coii- 

 siderable extent without affecting the results, but with coconut 

 oil the time given should be adhered to. 



