through the Medium of Stone. 



331 



was the best tliat circumstances allowed, as it was superinduced 

 on the old plan, without interfering with it. 1 should add 

 that, from the time of its adoption to this hour, the fires have 



never been lighted. 



In order to heat my hot-houses on this plan, I had the foot- 

 paths taken up; flag-stones {j^g. 84. a) were laid below: 



84 



upon them, on each side, was placed a line of bricks (b h) ; 

 upon these bricks were laid other flag-stones (c); and the 

 steam is admitted in the space between [d). In order to ren- 

 der the work firmer, it is desirable to have upright stones [e e) 

 against the sides of the flue, which prevent any shrinking in 

 the brickwork. The old fire flue {/) is remaining beside the 

 new one, in case the steam should not answer. The dimen- 

 sions vary a little, according to the breadth of the walks 

 round ; but the steam introduced at the bottom is allowed to 

 run thence round the house. The only inconvenience I 

 experienced in this plan was, that the steam, being introduced 

 all in one place, rushed forward with great force there, and 

 heated the house to a violent degree, but did not diffuse itself 

 with sufficient equality over the whole of tbe front walk. I 

 therefore got some gas-pipes in short pieces, made to screw 

 together : these were connected to the main pipe. {Jig 85.) 

 Into the openings [a b c) the gas- 

 pipes were screwed. One of these 

 pipes was a quarter of the length of 

 the front walk, the second was half, 

 and the third three fourths of the 

 length. The steam was thus deli- 

 vered at these respective distances, 

 and the flue under the walk was equally heated. It is further 

 to be observed, that the bottom of the tan-bed {fg. S^.f) is 

 also hollow, and that steam can be admitted underneath when 



