86 PROGRESS IN TURBINE SHIP PROPULSION. 



I do not quite follow Mr. Higgins' thought in saying that should an open feed-water 

 heater be employed there must be an additional pump. In either case there must be a pump 

 to remove the condensate from the condenser. This can as easily deliver the water to the 

 overhead open heater as to the feed tank, wherever it may be placed, in addition to which 

 there must be the boiler feed pumps, none other in either case being necessary. The open- 

 type feed- water heater is, of course, on the suction side of the feed pumps, and must, for 

 the reasons stated, be elevated above the feed pumps. 



Concerning the relative feed-water temperatures that may be obtained with the two types 

 of heaters, this is determined entirely by the quantity and the pressure of auxiliary exhaust 

 steam, which latter may be just as high in one case as in the other, with an advantage in 

 favor of the open type of heater, by reason of the intimate mixing of steam and water and 

 no dependence having to be placed upon the rate of conductivity through walls of tubes. 



It is important to exclude oxygen, as far as possible, from the system in order to avoid 

 corrosion. This plainly may be more completely carried out in the open type of heater by 

 reason of the feed water being very close to the temperature of evaporation. With the open 

 heater the pressure may be 5 pounds gauge and the water at 225° F., for example, and is 

 within 3 degrees of evaporating, while with the closed heater the pressure of the water might 

 be 300 degrees when at the same temperature it would be 200 degrees away from evapo- 

 rating, under which conditions but Httle of the suspended air would be expelled. 



Referring to Mr. Akimoff's remarks, it is not apparent what effect the number of bearings 

 could have on the torsional vibration except perhaps one of slight dampening. In the case 

 of the geared turbine the torque is so uniform that any variation therefrom would he of 

 such high periodicity that any response from the torsional elasticity of the shaft would hardly 

 be expected. Errors of balance, while productive of transverse vibrations, would not seem 

 to be productive of torsional vibration. There is always the possibility of torsional vibration. 

 the result of action of the propeller, but I have observed no gear trouble that I have sus- 

 pected as originating from this source. 



Replying to Mr. E. H. B. Anderson, I would state that he entirely misconstrues the 

 paper if he concludes its purpose is to show that success is only to be secured with gears in- 

 volving a patented device which is therein described. This is quite absurd, for rigid gearing has 

 been successfully operated for centuries, and no advocate of the floating frame type of gear has 

 ever suggested that any gears without this patented device could not be a success. The paper 

 does, however, endeavor to point out that pinions of floating frame gears may be safely sub- 

 jected to higher loads than those in fixed bearing gears when deflections and misalignments 

 incidental to frail hulls must be contended with. 



I think there can be no doubt but that the present great expansion in the adoption of 

 high-speed gearing, whether for shipboard or land use, was pioneered by Melville and Mac- 

 alpine. Their first large reduction gear was tested in 1909, the patents having been applied 

 for in 1907. The results of their work was published quite generally and was common 

 knowledge both in Europe and the United States of America at the time of the Vespasian in 

 1910. Sir Charles A. Parsons had experimented with gearing previous to this, but we learn 

 from the Transactions of the Institute of Engineers and Shipbuilders in Scotland, vol. 44, 

 page 216, that he had given up hope of success. 



The curve of tooth pressures employed for floating frame gears, given in the paper, ex- 

 presses a law which we believe to be quite well established, and I see no' reason why it is 

 not entirely comparable with the data which forms the curve for rigid bearing gears. The 



