THE TANK MODE OF HEATING. 



195 



pipe, and another takes it off when it is 

 necessary either to change the water or 

 to withdraw it entirely. The plants are 

 grown in large pots, kept about 4 inches 

 above the water in the trough ; the 

 branches being trained to a trellis, and 

 the fruit allowed to hang suspended. 



Mr Green also plants in a bed formed 

 in the following manner : over the tank 

 are laid strong stakes to support rough 

 boarding ; but this is so arranged that a 

 space is left open, both at front and back 

 of the bed, for the purpose of allowing 

 the vapour to ascend. It will be seen by 

 this description that atmospheric heat is 

 obtained from the small hot-water pipes, 

 and the bottom heat from the tank be- 

 low. This is one of the earliest examples 

 of the tank mode of heating, and would 

 be very complete if, instead of boarding, 

 slates or thin pavement were used, with 

 apertures in them for the ascent of vapour. 

 Wood is not only a bad conductor of 

 heat, but liable to decay, and this might 

 happen at a time when the plants were 

 in full bearing. In the annexed dia- 

 gram, fig. 249, the pipe for supplying the 



Fig. 249. 



tanks is shown at a, and the other for 

 withdrawing the water from them at b ; 

 c c are hot-water pipes for heating the 

 atmosphere ; d the bed in which the 

 plants are set ; e the trellis to which they 

 are trained. 



Excellent as the tank system of heat- 

 ing is, it is rather too much to expect 

 that it is sufficient to heat the atmosphere 

 of even a moderate-sized pit, far less 

 that of a large pine-stove ; although it 

 may afford quite sufficient heat for the 

 roots of the plants. The provisions made 



both by Mr Rendle, and also by ourselves, 

 for admitting a portion of heat from the 

 tanks into the atmosphere of the house, 

 are only well so far. It will be found 

 necessary in all cases, particularly upon 

 a large scale, to have means for heating 

 the atmosphere irrespective of the tank ; 

 and this can readily be effected in various 

 ways, such as by pipes connected directly 

 with the boiler, and used for heating the at- 

 mosphere, while the 

 tank heat should be 

 employed for the 

 bottom heat only. 

 This can be regu- 

 lated by properly 

 constructed stop- 

 cocks in the pipes, 

 fig. 250, and also 

 by sluices, fig. 251, placed in the tanks. 



Fig. 251. 



The mode of heating we have adopt- 

 ed at Dalkeith, wherever the tank is 

 employed, is to have a secondary supply 

 from pipes or flues, as may be most con- 

 venient. The tanks, of themselves, were 

 they even powerful enough to give atmo- 

 spheric as well as bottom heat, would, un- 

 less exceedingly well jointed and kept in 

 repair, give out too much humidity for 

 plants in general to thrive in, unless it 

 be tropical plants and orchids. It may 

 often happen with regard to a house 

 heated by a tank, that at some particular 

 season a dry atmosphere may be required, 

 such, for example, as the prolongation of a 

 late crop of grapes. This could never be 

 obtained from the tank alone ; but with 

 pipes attached to the boiler, irrespective 

 of the tank, such could be got with the 

 greatest ease— namely, by shutting off the 



