[mellanby] EFFICEENCIES OF MULTIPLE-EXPANSION ENGINES 131 



to each cylinder. The indicator cocks were screwed directly into the 

 cylinder bosses, thus avoiding any intervening pipes and cocks, and the 

 inaccuracies almost invariably connected with them, due to insufficient 

 area, and condensation of the steam between the cylinder and the 

 indicator. 



Crosby indicators were used on all the cylinders except on the low 

 pressure, where Tabor indicators were used. One man was stationed at 

 the indicators on each cylinder, and when the signal was given the cards 

 from all the cylinders were taken as nearly as possible simultaneousl3\ 

 It was the duty of each of these observers, as soon as the cards were 

 taken, to read the thermometer and pressure gauge on the steam chest of 

 the cylinder at which he was working. One observer was told oft' to read 

 every five minutes the gauge and thermometer on the steam pipe before 

 the first cylinder, the vacuum gauge and thermometer on the exhaust 

 pipe, and the vaciium gauges on the condenser. 



A tachometer which recorded the number of revolutions per minute 

 at which the engine was running was attached near to one of the brakes. 

 One man was stationed at the brake, whose duty it was to keep the 

 speed of the engine constant by increasing or decreasing the load on the 

 dynamometer as the tachometer showed the speed to be going above or 

 below what was required. When any alteration of weight was made, 

 the amount of it, and the time of alteration were noted down, so that the 

 accurate average B.H.P. throughout the trial could be obtained. When 

 in addition to the water brakes the rope bi'ake was used, the weight 

 shown by it was read every minute. The condensed jacket steam was 

 allowed to run into tin vessels, and its amount obtained by weighing. In 

 the early trials it ran into small tins each capable of holding two lbs. 

 Two observers were stationed to look after it. and as soon as each tin filled, 

 one observer signalled to the other, when, the cylinder it came from, and 

 whether from the top, bottom or sides, as well as the time of filling, was 

 noted down. In the later trials this method of measurement was 

 abandoned. The water was allowed to run into larger vessels, and 

 the amount that had flowed into them was weighed at the end of the 

 trial. The former method has, however, the advantage that the rate of 

 jacket drainage can be observed frequently and maintained uniformlv at 

 its best. 



At the beginning of the session 16 preliminary compound trials were 

 made, chiefly with the object of allowing the 11 observers, who were 

 composed of the fourth year class in mechanical engineering, to* become 

 accustomed to the method of making the observations. 



Description of the Trials. 



The notation used in designating the trials is that used by Willans 

 in his papers on Engine Trials, read before the Civil Engineers' Institute. 

 Thus T 155-123 No. 26, indicates that the engine was working as a triple 



