LXVIII. DISCUSSION. 
several years experience with the self contained simple and duplex 
machines, they have now produced a triple expansion horizontal 
surface condensing steam engine combined with three separate 
compound ammonia compressors arranged all on the same bed 
plate ; the three steam cylinders are placed parallel to each other 
and are connected up toa three throw crank shaft ; the ammonia 
cylinders are placed tandem to the steam cylinders, and are driven 
by prolongations of the piston rods; most of the parts are inter- 
changeable, and the steam engine can be worked either single, 
compound, or triple expanion, condensing or non-condensing as 
the case may be. The crank shaft is also in three separate parts 
connected by two couplings. 
Mr. Statham was surprised at Mr. Cruickshank’s remark 
regarding the insufficiency of duplication in marine plants, as it 
has been the practice of many makers for years past to supply 
duplex machines on board ship. Marine plants should have 
always a duplex machine and duplicate coolers, plants of this 
type have been taking home valuable cargoes of meat through 
the tropics using only one half of the machine, the other half 
being kept as a stand by, and generally only used when the ship 
is loading up, and it is necessary to cool the holds down rapidly. 
Mr. Hovuauton said the author’s statement as to the part play ed 
by Australia in the early history and development of the freezing 
machine was most valuable, as it placed on record the names of 
the men to whom credit was due. In the discussion on the first 
paper read on refrigerating machines before the Institution of 
Civil Engineers,! a long list of the works bearing on the subject 
which had appeared up to 1874 was given, and it shewed how 
many men had been at work trying to perfect the machine at that 
early date. The very low temperatures attained in the expansion 
cylinders of machines using air for refrigeration affected the 
strength of the steel piston rods and they broke, although the 
stress was very much within the ordinary elastic limit of the 
Vier 
1 Proceedings Inst. C.E., Vol. xxxvil., p. 271. 
