EXPERIMENT ON HEATING EARTH. 187 



found the plants coming up in all directions, blanched of course 

 by the leaves which had become "wet by rains, and even perhaps 

 heated slightly." From this cause the ChamceropSf Blandfordia 

 and some evergreen bulbs had lost their foliage, but were alive 

 and preparing for fresh growth, and the little axillary bulbs of 

 Begonia Martiana^ and discolor were coming up in numbers 

 around the old plants. 



The leaves were replaced more loosely, some sea-kail pots 

 being first put over a few of the plants. And with this terminates 

 my report for the present. 



I have endeavoured to give a truthful and unvarnished account 

 of a very rough experiment. It is poor in results, but the expe- 

 rience of one season can do but little to elucidate a principle the 

 probable developments of which are almost infinite. The huge 

 and rapid growth of many of the finest tropical plants, over- 

 powering others of slighter constitution, by their shade and the 

 exhausting action of their roots, will set bounds to the gro"wing of 

 any great number or variety of plants in the small spaces that 

 considerations of economy would allow us to heat artificially. A 

 profusion of water was reijuired at all times. A great draw- 

 back was the destructive effect of strong winds upon the foliage, 

 and the perpetual *' struggle for existence " going on among so 

 many plants of differing constitution, habit, and strength. In 

 the hope that some of my brother horticulturists may take up the 

 subject during the coming season, I may venture upon a few 

 hints and suggestions, derived from my own experience. A brick 

 partition down the centre of the bed in the direction of its length, 

 would enable us to isolate the strong growing sorts in loamy 

 soil in the background, while the front strip, or partition, might 

 be filled with compost suitable to the more delicately rooting 

 herbaceous subjects. This again might be subdivided roughly by 

 a slate or two, so as still further to vary the characters of the 

 soil, as, for instance, where the use of peat might be specially 

 indicated. There would be no difficulty or oomphcation in such 

 an arrangement. The top of the main partition should be just 

 flush with the surface of the soil, so as to be hidden, for 

 appearance-sake, with a little sprinkling of earth. The heating 

 power of a 3-inch pipe, applied as described, proved amply suffi- 

 cient ; and I have no doubt but that any length of border might 

 be effectually heated in the same way, provided that the width 

 and depth of the mass of soil were the same. The proposed 



