NEW DEVELOPMENTS IN HIGH VACUUM APPARATUS. 187 



The curves upon Plate 08 would be very interesting if correct and authoritative. If 

 iliey were determined by experiment, it would be interesting to see the determining data 

 and observations; as a matter of fact, within the week I have had estimates given me by 

 an engineer whose experience and opportunities for obtaining such data were at least as 

 good as that of any engineer in the country. On the smaller sizes his allowances run from 

 50 to 100 per cent greater than shown on Plate 68. Mr. R. J. Kaula presented to one of 

 the British societies a curve of leakage as reproduced in London Engineering of April 18, 

 1019, in which, for 50,000 pounds, he shows 41 pounds, and for 100,000 pounds of steam, 

 08 pounds of air per hour, as against 20 pounds and 32.5 pounds respectively as shown on 

 Plate 68. 



It will be interesting to state in this connection that the 30,000-kilowatt units installed 

 at the water side station of the New York Edison Company (approximately .350,000 pounds 

 of steam) are readily maintained in service at between 16 and 23 pounds of air leakage per 

 hour, an amount which corresponds upon Plate 68 to a steam consumption of from 32,000 

 to 62,000 pounds per hour. 



The New York Edison Company measures once a week, through a gasometer, the non- 

 condensible vapors discharged from each condensing unit; when the leakage increases to, 

 say, twice the above amount, the vacuum is so much affected that the unit is overhauled, at 

 which time the leakage is brought as low as 13 to 15 pounds of air per hour, or at special 

 times even lower. Other large stations similarly equipped report practically the same result. 

 A complete discussion of this subject would, make a lengthy paper, but evidently the curves 

 on Plate 68 are hardly capable of miscellaneous application. 



The separate wet and dry system referred to by the author has been used by Worthing- 

 ton for many years in stationary practice and has been extended to include vertical marine 

 pumps of the twin type. The steam ejector has the important advantage for marine work 

 of reduced space and weight, but many disadvantages and complications, as may be seen 

 from the description and layouts accompanying the paper. The Worthington Company have 

 carried on investigations and experiments in the ejector field for many years and are now 

 offering ejectors either of the condensing or non-condensing type. The paper refers espe- 

 cially to the adoption of the air ejector during recent years and conveys the impression that 

 this has been to the exclusion of the twin or twinplex type. The adoption of the ejector 

 would have been a slow process except for the unusual conditions brought about by the war. 

 The battleships, cruisers, and destroyers of the United States Navy were equipped with the 

 twinplex type of pump (upwards of 400 during this period for destroyers alone), and in 

 a few cases, where the ejector was installed by the Navy, it was only as an auxiliary or in 

 conjunction with a pump of the twinplex type. 



The Worthington Company's capacity and production were greatly increased during the 

 war to care for the government requirements, and it was impossible to produce the additional 

 vacuum pumps which would have been required by the cargo vessels, and equally impossible 

 under those conditions to undertake the further development and manufacture of an additional 

 line, such as ejectors. 



The reasons for the building of the cargo steamers and the essential features of design 

 were quite different from those either in the ordinary merchant or naval service, and a sim- 

 ple, light, cheap apparatus that could be quickly produced and installed was therefore per- 

 missible. 



This is no place to introduce a discussion of the Worthington ejector, but I will state 



