By Sir George Stewart Mackenzie, Bart. 175 



range. It is therefore no disadvantage to the globular form, 

 that it does not admit of being extended more than thirty 

 feet. I can imagine that the inside of such a house must be 

 very beautiful, when covered with clusters of ripe grapes, 

 and that the effect will far exceed that tame uniformity 

 which is seen in houses of ordinary construction. 



On referring to the plan, Fig. 2, you will perceive, that the 

 entrance does not interfere with the glass, which is immove- 

 able. From K to L, the wall supporting the glass is arched, 

 and another wall is built, five feet behind, in which windows, 

 or openings, of any convenient form, for ventilating the 

 house, may be made. For ordinary airing, there are open- 

 ings in the low front wall A A, and others at the top of the 

 back wall, as seen at B, in the section Fig. 3. By means of 

 these, a current of air may be conducted along the inside 

 of the glass. It is by no means impracticable to make por- 

 tions of the glass moveable, should that appear to be neces- 

 sary. The plan for the back part may be varied, without 

 interfering at all with the glass. 



I cannot too strongly recommend that form of flue which 

 I have given in the Memoirs of the Caledonian Horticul- 

 tural Society, and which I call the embrasure flue. I have 

 compared its effects with that of an ordinary flue, and have 

 found, that the heat is much more easily kept up than in a 

 common one, with a saving of more than one-fourth of fuel. 

 An inspection of Fig. 4, will satisfy you, that a greater 

 heated surface is exposed, and that the heated air must pass 

 much more slowly in this form than any other, while at the 

 same time there is a sufficient draught. 



The ribs of the semi-dome, which will form the astragals 



