Retrospect he Criticism. 377 



I conceived that the siphon would also circulate hot fluids on the 

 same principle of density or condensation ; and also that it would cause 

 more rapid circulation, which appeared to be a great desideratum. I there- 

 fore caused two tin vessels to be made, and connected them near their 

 bottoms with a straight tube. I tried the experiment with a siphon, first 

 about 3 ft., then 10 ft., and afterwards 20 ft. above the surface of the water 

 in the vessels ; and I found that the circulation was more or less rapid 

 in proportion to the elevation, and the means taken to increase or diminish 

 the difference of temperature in the ascending and descending legs. One 

 insu]5erable objection, however, which I then conceived would for ever 

 operate against its utility (namely, the diminution of the boiling point of 

 water in proportion to the elevation) caused me to suspend my operations 

 for some time ; but still meditating on those experiments, and occasionally 

 trying the effect of my 20 ft. siphon, I found that a low temperature of the 

 water would cause a difference of level between the two vessels of 1 in,, H in., 

 and even 2 in., if the connecting tube were stopped ; and thus a circulating 

 current was evidently created with the force of this fall. I quickly saw 

 that this property might, indeed, be usefully employed to circulate boiling 

 fluids to almost any extent, and with great rapidity ; and even if required, in 

 open channels on the ground, by using one siphon only, to take up the fluid 

 and give it a fresh impulse after it had again passed through one of a series 

 of boilers. It was also evident that, by lengthening or extending the sur- 

 face of the tube from this last boiler, the fluid might be suffered to cool so 

 much as might be needful before it again ascended in the siphon, &c. &c. 

 But as all this, and much more, is clearly described in my specification and 

 pamphlet, I need not enlarge on it here ; and shall only remark, that, after 

 much experience and application, I am decidedly of opinion that the 

 various modifications of which my plan is susceptible will, in general, ulti- 

 mately supersede every other plan yet known, as I am well assured of its 

 very superior powers and efficacy, united to perfect freedom from all dangex' 

 in any case, 



I certainly did not expect, in the article alluded to in your Magazine, 

 that you could enter into all the details of my plan, or that a hasty opinion 

 formed of it could be perfect ; but what was there said was honestly said, 

 and I feel in the highest degree grateful to you for it : its real merits, if any, 

 must be left for the experience of posterity to decide on. 



You may now easily conceive what my feelings were when I received a 

 letter from you in the fall of 1830, saying that " Mr. Kewley had made great 

 improvement in the siphon mode of heating, which would appear in the 

 next Number of the Magazine." [See Vol. VI. p. 377.] When this Num- 

 ber appeared, I found that his improvement was the substituting a pump 

 for the filling cock. Surely it cannot be said that a pump was never before 

 employed to fill a tube, or that a pump is at all essential in the action of the 

 machine patented to me. My specification says, " the invention consists in 

 causing water, oil, or other fluids, to circulate through the medium of a 

 bent tube by raising the temperature of the fluid at one end of the tube, so 

 that it is always warmer than that at the other;" and I then proceed to 

 show how that tube which I call a thermosiphon may be filled, I adopted 

 what you call the cock system, in preference to the pump, as I knew the 

 mode of filling with the cocks and air-plugs was perfect, and that cases 

 could be proposed where the pump would fail to extract all the air, which, 

 of consequence, would have had some tendency to vitiate the patent, had I 

 adopted the pump in the specification. At any rate, one of the effectual 

 modes of filling and refilling the tube, out of the many plans that might be 

 devised, was sufficient for the purpose of the specification, and I adopted 

 that which appeared to me to be the most perfect. Mi-. Kewley's substitu- 

 tion of the pump is, therefore, a palpable infringement; and if you reperuse 

 or recollect the article in the Magazine now "alluded to, I am sure your 



