NEW DEVELOPMENTS JN HIGH VACUUM APPARATUS. 193 



and look for tlie leaks and make them tight until you get the required vacuum. We cannot 

 say exactly what the air leakage is going to be, and must provide a reasonable excess of ca- 

 pacity. It is thought marine engineers should get together and decide on what are reasonable 

 air leakages for different types of plants. I think, however, that Mr. Kothny's curves are 

 useful and believe they are safe for designers to use. 



The illustration on Plate 77 shows a comparison between the piston type pump and the 

 air ejector. I have not checked this comparison, but, offhand, it looks about the same as we 

 laid out in our ships. 



In regard to the complexity of the installation,, we have not found it specially complex. 

 The feed tank which was mentioned as being somewhat special is really not special at all. It 

 merely has two pipes projected into it, with perforations. That is the only special feature 

 about it, except the re-circulating pipe which is connected on to it to carry the water back 

 to the condenser for cooling. 



Attention is called to Plate 80, showing an air ejector installation for a steam-engine 

 vessel. I do not believe it is generally understood that air ejectors can be used to advan- 

 tage for steam-engines. I have not made an exact comparison of the two types for engine 

 vessels, but I would say, offhand, that there was not any disadvantage in using air ejectors. 

 For most cargo steamers I think it is desirable to leave off the automatic thermostatic control, 

 although that is something that does not do any harm, and may do some good. 



We have turned out twenty-seven vessels with the C. H. Wheeler "Radojet" installa- 

 tion, and in all cases these vessels have given excellent results. The condensing plant was 

 carefully designed all the way through. The specifications call for a 38-inch vacuum with a 

 sea-water temperature of 70 degrees, and in practically all cases, after the leaks have been 

 taken care of properly, we have obtained from ^4 to >4 inch better vacuum in service than 

 we designed the installation for, which is considered a good perfonnance. 



In regard to the steam consumption of the piston pump, and the air ejector, it would 

 be interesting to have accurate figures on the two steam consumptions. We made a com- 

 parison of these values in selecting the air pumps for our vessels, and, as I recall it, the 

 Radojet was less than the piston pump, although I would not be positive. We found that 

 the Radojet was much lighter, and we considered it simpler. It was more compact, and we 

 considered that it gave us a better engine-room arrangement — more floor space and more 

 room around the engine. 



There is one feature in connection with the use of these air ejectors for turbine-driven 

 vessels which should be taken care of, and that is the turbine drainage. The turbine drain- 

 age is of equal importance with the lubrication for the safe operation for turbines. Great 

 care must be taken to' provide a safe drainage for the turbine, so that the water will not back 

 up and interfere with the blades and cause stripping. In some of the vessels which we built 

 we installed steanr ejectors; in some small piston pumps on the tank top, and in others the 

 two together. We installed both an ejector and a steam pump and provided suitable means for 

 observing the water in a drain well, so that the operator could keep track of it. I want to 

 emphasize the importance of providing for absolutely safe drainage where these air ejector.-? 

 are used. 



In regard to the heat balance and auxiliary exhaust steam, etc., in some cases you can 

 control this in your design and in other cases it gets away from you. In most cases, in or- 

 dinary cargo ships the auxiliary exhaust is all condensed in the feed heater. In most turbine 

 vessels you have more auxiliary exhaust than the feed heater will take care of. The surplus 



