37S 



NATURE 



[February 14, 1895 



convenient foini of apparatus for determinations by this method. 

 It consists of an iron vessel about a foot in fiiame'er, furnished 

 with a loose cover ; this forms the condenser. A small pipe and 

 cock in the steam-pipe deliver steam through a small orifice 

 nfar the steam-pipe into arothir pipe, throi'gh which it passes 

 itto the condensing water. An .igitator and a >on-iive ther- 

 momeier are provided in the condenser. F< r weighing the 

 amount of steam condensed, ihe whole condenser is suspended 

 from a h)dro tat, «hich permits extremely accuraie deiermina- 

 tion of any chat ge of weight. The hydrcstai i> ba anced by 

 weights till the pjinter is at a fixed mark b; fore and alter con- 

 densing the steam. The condcn'-ing mi tliod, the author said, 

 is .'tri tly accmate in principle, but dilTicult to larry out in a 

 satisfactory manner. In order toovtrcoire the e ditficuhies a 

 method oi cortmuius condensation h,Ts been in ruducid. By 

 this steam and ci>ld tkatcr are both supplied a' a constant 

 rate, and the condenser acquires a steady tempnature, which 

 can be very accurately observed. .\ diri^ram of a continuous 

 injection condenser was shown en the wall of the tl eatre. Steam 

 passes from the stcam-pif e to a small injector. The condensing 

 water is drawn from a tsnl:, and the mixed water andomd nse.l 

 steam are discharged irto another tank. The two tanks are 

 placed on platform weighing machines. Two thermometers give 

 the tempi rature of the condensing water (the water used for 

 condensing the steam), and of the mixture of condensed steam 

 and condensing water. The difference of the tclal weight in 

 the two tanks, alter any interval ol time, is the sicim condensed 

 in that time. 



K superhealing method, which was introduced about the year 

 ■ S90, by Mr. G. Uarrus, was next referred to by the author. 

 The s'lam to be ie-t«d is passed ihioogh an inner chamber 

 jacketed by superheated steam. The sample of steam lo be 

 tested wa< thus dried and supeiheatid at the expense of heat 

 borrowed from the jacket. To avoid measuiing the steam, an 

 attempt was made to secure that equal weights oi steam passed 

 through the inner chamber and through the jacl'Ct. In th.it 

 case the wetness of the steam can be calculated from observa- 

 tion of the Icmperalutes only. The method, the author said, is 

 accurate in principle, but appears to be difficult to carry out 

 satisfactorily. 



The wire drawing calorimeter was next descr.bed. This and 

 the separating method the author considers most neatly fulfilled 

 the necessary conditions requisite for iiieasuirng the quamity of 

 moisture contained in steam. This c-ilorimctcr consists of 

 two chambers, steam passing from the first lothc second through 

 an aperture Vii inch in diameter. The full steam ]>ressurc is in 

 the hist chamber, and the pressure in the second diflers litl'e 

 from atmospheric pns-ure. Thermometers give the tcmpeia- 

 ture% in the chambers (which are protected from radiation), and in 

 this way the quantity ol water present in the steam isesiimaled. 

 The steam is allowed to flow through ihe app.iratus for twenty 

 minutes or more, when tVc ictrperaturcs become nearly steady. 

 No weighing is required, and temperatures only have to be 

 observed. The obscrvaiions can be continued as long asdcsircd, 

 so as to otita n a mem value for the dryness fraction rom a con- 

 siderable quantity of steam. If the steam is very wet, the tem- 

 perature in the second chamber falls to about 212 , showing th«t 

 wire-drawing to atmospheric pressure is insufficicrit to dry the 

 ileam. Practically the instrument cannot be used if the wet- 

 ness e»ceedi the valuci given in the following table, Ihe 

 presiufcs being in lbs. per square inch, and the atmospheiic 

 pressure being assumed at \\"j lbs. 



Initial [rciture loiiial preft«iirc Initial temr«ra- 

 (atxolulc). (gau^e). tura !•. 



29V • IS'2 - 250° 



672 525 ... 300° 



«35'' • • '234 •• 35"° 



2477 2330 ... 400° 



Two conrlit ions arc nrccs ary for accuracy in using this method. 

 The second chamber must be large enough for the eddies lo die 

 out be'orc the steam leaves ihe chamber. Radiation must be so 

 f.ir prevented that the steam in the chamber is not sensibly 

 cooled. A calorimeter by which the separating and wirc- 

 ilrawing mithoiU were combined was also explained ly the 

 author. This ii the globe calorimeter which is a well-arranged 

 apparntu*. 



The Camming lapethcating method was also described. A 

 vessel is filled wiih the steam lo Ire tested, and then heated by 

 a jac'iel. As it is healed, the rise of pressure in the inner 

 vesiel is oViscrvcd, the volume being conrlant. So long as the 



steam is moisi, the pressure will rise with the temperature 

 according to Ihe law lor saturatcel steam. The moment all the 

 moisture is evaporated, the r.ite of rise of pressure with tem- 

 perature will become much slower. 



The well-known salt test was next alluded to by the author. 

 This, however, he pronounced lo be incorveniert and unlrust- 

 worihy, excepting perhaps in the case of .\ boiler subject to 

 marked priming action. 



The general conclusion drawn by the author was that the 

 wire-drawing calorimeter without separator is the most con- 

 venient .ind accurate for steam with less than about 2 per cent, 

 of moisture. For steam containing more moistuie, the separat- 

 ing calorimeter without wire di awing apparatus is accurate 

 enough and convenient. The use ef the separator and wire- 

 drawing calorimeter combined is more troublesome, especially 

 if, as is desirable, a condenser is also used lo determine the 

 amount of steam passing through the separator. In cases 

 where there is much priming, it would seem best to take the 

 whole of the steam ihroui^h an ordinary steam-separator, 

 measuring ilie amount ef water trapped, and then to test by a 

 wire-elrawing or separating calorimeter the dryness of the steam 

 after passing the separator. With priming much of the water 

 probably flows along the bottom of the pipe, ami it appears 

 impossible that a sample can be obtained containing an average 

 proportion of steam and water. It is recommended by Prof. 

 Carpenter that the sample of steam to be tested should always 

 be taken from a vertical, not from a horizontal steam-pipe. No 

 doubt there is rather more tendency for water lo flow along the 

 bottom of a horizontal pipe than down the sides of a vertical 

 pipe; but merely taking steam from a vertical pipe does not 

 ensure freedom from error, especially if the amount of moisture 

 in ihe steam is considerable. Variations in tests for wetness 

 are doubtless often due lo the difficulty of getting a true average 

 sample of steam : and it would seem that errors are generally 

 in the direction of under-estimating tli'^ amount of moisture. 



A long and interesting discussion, which was adjrurned from 

 Ihe Thursday until the Friday evening, was held on Prof. 

 I'nwin's paper. The chief point touched upon was the nr.ethod 

 lo be adopted in gelling a true sample of steam for analysis. 

 This undoubtedly is the great difliculty that has to be overcome 

 before a satisfactory method of determining the amount of 

 moiilure in steam can be arrivid at. The majmity of speakers 

 were of opinion iliat water entrained in the steam would hang 

 to the sides of Ihe pipe, and a gooel many suggestions were 

 made with a view to shifting the collecting nozzle over the whole 

 area of the cross-section of the pipe, or else lo give such an 

 orifice to the nozzle as would cover a large part of the pipe 

 area. 



In his reply to the discussion, Prof. Unwin explained that 

 this did not seem to him Ihe true light in which ihe problem 

 should be regarded. With steam rushing through a pipe at 

 hit;h speed, eddies would I e set up which would be sufiicient to 

 thoroughly mix the sleam and water so that there would be a 

 fairly homogeneous mixture. The Iruc difficulty arose from 

 the checking of the velocity of the steam at the collecting 

 orifice, an action which resulted in water accumulating so that 

 an excess of moisture was shown in the sainirle drawn ofT". In 

 order to overcome this, he had devised a collecting nozzle con- 

 sisting of a bent-over tapered pipe, Ihe orifice of which was at 

 the small end, and was pointetl towards Ihe flow of steam. By 

 adopting the necessary dimensions for the collecting nozzle, the 

 steam collected would not be checKcd in velocity at the cedlect- 

 ing orifice, and therefore moisture would not be deposited at 

 that point. 



Captain Sarkey's paper was ore of corsiderabic length, and 

 alil ough dealing with one 1 oini only of engine design, was of 

 great interest '■ engineers. It was illustrated by a large num- 

 ber of diagrams hung on the wall of the theatre. Without these 

 it would b; extremely diflicult 'o give a fair idea of the course 

 of reatoning followed by the aulfor in discussing the nrerits of 

 Ihe two systems of governing engines. The pa) er, as Ihe author 

 said, was an elaboration of one sect Hn of a paper contributed lo 

 the Institution of Civil Rnginecrs by the late .Mr. Willans. 



.Speaking broadly, it may be said that the author's opinion 

 was lira', the popular verdict in favour of variable expansion 

 governing may for many purposes be accepted, yet its advan- 

 lages were commonly much overrated, and in some cases it had 

 no aelvantage at all. 



It would be impossible within Ihe limits of our report to trace 

 out the respective merits of Ihe use of Ihe throttle valve and 



NO. 



1320, VOL. 51] 



