50 



which is spread in a thin film over both the corrugated sur- 

 faces by fine holes in the funnel shaped cover D. 



The bottom of the space K is covered with a ring Z, 

 which lets the milk out by the faucet g. 



This heater approaches in effectiveness one suggested by 

 me in 1894 for cooler, namely by having two corrugated 

 cylinders made with spiral corrugations so that the inner 

 t>ne could be screwed down in position, leaving only half or 

 one-quarter inch space for the milk. I believe it would be the 

 ideal heater (and cooler) where no motive power is to be used. 

 The milk would hae to follow the corrugations somewhat as 

 shown in Fig. 43. I once corresponded with manufacturers 

 of copper expansion joints for large steam pipes, but the price 

 seemed too high to make it practical. 



A very simple device is shown in Fig. 48. Mr. Bentley, of 

 44 Ohio, used it and I have also had one 



made, but with several sets of cans 

 within each other. It is really an 

 adaption of a cooler illustrated in Dr. 

 Fleischmann's book, invented by Jell- 

 inek Romanowsky years ago. A tube 

 H is soldered in the center of a shot- 

 gun can AAAA, having holes in the 

 rim below. This is placed in a larger 

 can BBBB. The latter is placed in a 

 tank with water kept warm with a 

 steam jet (or cold with ice if used as a 

 cooler). The can AA is also filled with 

 water, warm or cold, and the milk en- 

 ters at H, flows through the holes in 

 the rim of the can AA, and passes up between that and BB 

 and out at M. 



In Sweden De Laval solved the protection problem 

 in a similar but more elaborate manner, as shown in Fig. 49. 



This apparatus consists of two closed double vessels fit- 

 ting one into the other in such a way as to form concentric- 

 narrow apertures of large surface, through which the milk is 

 forced. The aperture is only about of an inch and the milk, 

 which is kept in constant motion, is rapidly and evenly heated, 

 without allowing any albumen to coagulate. 



Each vessel has a pipe (a) which passes down close to the 



H 



5i*x 



Fig. 48. 



