1842.] 



THE CIVIL ENGINEER AND ARCHITECTS JOURNAL. 



193 



as their authority is invalid ; and it would be in every sense super- 

 fluous to address ourselves to them. With such we have no concern : 

 Dor is the Polytechnic Society anything more to us than Hecuba was 

 to the Prince of Denmark. 



But farther ; the mere surface area of the spike is not the measure 

 of transmissive power. What then is the measure ? or is there a 

 measure at all ? If amid all the riot and pretension we have witnessed, 

 it should turn out that there is no measure of the beneficial effect of 

 the spikes at all, is it possible to resist the conviction that the whole 

 affair is a pitiful juggle, the effect of which is to delude the ignorant 

 and unwary ? And if there be a measure, why are we not told what 

 it is? Again, it is said the transmissive power of the spikes increases 

 with the length, but does not proceed beyond a certain limit, and 

 three inches is the best length. But all this is unsupported assertion 

 merely : the question is, mhy, as far as theory is concerned, there should 

 be a limit to the beneficial length of the spikes. The spikes, we are 

 told, operate in promoting evaporation by virtue of their surface — 

 their heat-receiving and absorbing surface, as it is called. A spike 

 three inches long is alleged to be better than a spike two inches long, 

 because it has more surface ; should not, then, a sis-inch spike be 

 better than a three-inch spike, for the same reasoa? It is no answer 

 to tell us that such is not the case. The question is, whij is it not the 

 case ? Why, if spike surface be beneficial at all, should twenty square 

 inches not be better than ten ? 



Turn we now to what our correspondent A. S. advances in favour 

 of the spike system. He begins by admitting the truth of the obser- 

 vation we made in our former article upon these projects, namely, 

 that the spike theory is incompatible with the received theory of 

 conduction, and that, therefore, one of the two must be wrong ; but he 

 adds that it is the received theory and all the world that is wrong, and 

 the spike theory and Mr. Williams that are right. His line of argument 

 is as follows : — 



The experiments of M. Delarive, says he, show that the conduct- 

 ing power of some fibrous substances varies exceedingly, according 

 as conduction takes place in a direction transverse or parallel to their 

 fibres. Wrought iron, being a fibrous substance, may fall under the 

 same law, so that heat may be able to penetrate much better in the 

 direction of the fibres, as in the longitudinal direction of a rod, than 

 across the laminae or fibres, as through the thickness of a boiler plate. 

 He further says that it has been found that increasing the dimensions 

 of a conducting substance in the direction of the conduction produces 

 only a slight retardation in the passage of the heat; a fact, he says, 

 sufficient to destroy all confidence in reasonings founded upon the 

 ordinary hypothesis : that, therefore, the power of transmitting heat 

 may depend on something more tlian mere difference of temperature; 

 and the spikes may, consequently, be beneficial. 



All this, though extremely hypothetical, is very specious; yet it 

 admits, we conceive, of a very satisfactory and obvious reply, if, as 

 A. S. assumes, heat penetrates with unusual facility in the direction of 

 the fibres of wrought iron, it might certainly be conceived to be de- 

 sirable to insert in the plate iron of the flue discs sliced from the end 

 of an iron bar, or for forming the flue of iron, the fibres of which lay 

 in the direction in which the heat was required to penetrate. But to 

 justify the use of the spikes, it is necessary to suppose not merely that 

 heat is transmitted more rapidly in the direction of the fibres than trans- 

 versely to them, but that heat is transmitted more rapidly through a 

 spike three inches long in the direction of the fibres, than through a 

 spike half an inch long in the direction of the fibres. If the doctrine 

 advocated by A. S. be correct, we may conceive that if all the spikes 

 of a boiler were cut off at the level of the surface of the plate, the 

 discs of iron left in the plate might render the flue surface more per- 

 vious to the heat, on account of the greater conducting power of the 

 iron in the direction of the fibres. But this increased permeability 

 would be impaired or extinguished, if the heat had to travel through 

 a distance of several inches of spike before it reached the plane of 

 the flue surface ; unless, as we have already stated, it be held that the 

 heat travels quicker through a very thick than through a very thin 

 piece of iron in the direction of the fibres, an idea too preposterous 

 to be entertained for a moment. Instead, therefore, of the ar- 

 gument of A. S. proving the advantage of the spikes, it rather 

 proves that if they be already inserted in any boiler they should be all 

 cut off. 



But further ; A. S. asserts that no regard is to be bad to the received 

 theory of conduction, because an increment in the dimensions of a 

 body in the direction of the conduction only slightly retards the pro- 

 pagation of heat; in other words, because the increment of dimension 

 and the retardation of conduction do not vary in the same ratio. But 

 what of that? The common doctrine of conduction does not specify 

 any particular ratio in which this variation shall necessarily take place. 

 It only says that heat is transmitted through a thin piece of iron more 



rapidly than through a thick piece of iron, and that more heat is com- 

 municated by a square inch surface of hot iron than by a square inch 

 surface of cold iron. In this sense it is rather to be accepted as a 

 self-evident truth than as a questionable hypothesis; and the en- 

 deavour to overturn this doctrine is plainly quite as hopeless as the 

 attempt would be to disprove the most obvious truth in the whole 

 range of our experience. Besides, if conduction do not vary as the 

 temperatures simply, it must vary as the temperatures and masses 

 conjointly. The quantity of heat, or what is equivalent, the quantity 

 of matter, in a body will then become an element of the rapidity with 

 which heat is communicated ; and the admission of this element 

 immediately throws us upon the " imperial absurdities" to which we 

 have already adverted. 



But perhaps the most extraordinary proof of the hollowness of 

 A. S's doctrine is the fact that he is himself, it would appear, unable 

 to listen to the consolation it affords. No sooner is the new doctrine 

 propounded than it is thrown overboard, as a thing of no worth or 

 moment, and our credibility is solicited in favour of another doctrine, 

 which, we are assured, is so much in accordance with the laws of 

 nature, that it ought rather to be adduced as an explanation thaa 

 adopted as a hypothesis. This doctrine is only a reproduction of that 

 analysed in our former remarks, as proceeding on the assumption that 

 there is a stratum of cool air in the flues resting upon a stratum of hot 

 air — a supposition which we showed was just as much in accordance 

 with the laws of nature as that a stratum of water should swim on 

 a stratum of oil. The hot air, impelled upwards by its buoyancy, 

 will always rise to the upper part of the flue, and as soon as it is no 

 longer hot from the transfer of its heat to the water of the boiler, it 

 will descend by the sides of the flue to the level proper to its tem- 

 perature, without any considerable intermingling of the particles of 

 the ascending and descending currents, as supposed by A. S., and 

 therefore ^vitliout any considerable tendency from that cause towards 

 an equalization of their respective temperatures. Nor can the rapidity 

 of the current of the hot air through the flues interfere with the eco- 

 nomy of these molecular movements ; the current through the flues 

 being merely a relative motion of the hot air and the boiler, and not 

 of the particles of the hot air with respect to one another. 



This explanation, then, of the rationality of the spike system en- 

 tirely fails; but even if it did not fail it would be nothing to the 

 purpose, for this doctrine of the mode of action of the spikes, whether 

 good or bad, is not the doctrine of Mr. Williams. It is not the thing 

 to which he has called upon us to yield our assent, and is, therefore, not 

 the thing into which it has been our present purpose to inquire. We 

 come back, tlien, to the very point from which we started. We have 

 no alternative but to admit that the received theory of conduction as 

 we have explained it is right, and it is equally inevitable that we 

 believe the spike theory to be wrong, and this gentleman's whole 

 project an ineffable absurdity. 



We have hitherto confined our remarks to theoretical considerations 

 only : were we to extend them to the practical bearings of tl'.e subject, 

 the scheme would appear more preposterous still. And first it might be 

 remarked, that all the spike theory profes^'en to accomplish is to diminish, 

 iu some measure, the quantity of flue surface requisite to produce a cer- 

 tain evaporative effect. It does not profess to save fuel — to aid combus- 

 tion — to increase the efficacy of the steam, or to diminish the expense 

 of maintenance. Supposing even, therefore, the spike theory to be 

 correct — supposing that a square foot of flue surface spiked were 

 superior in evaporative efficacy to a square foot of flue surface un- 

 spiked, we do not believe that any practical man, whose opinion is 

 entitled to respect, would recommend the adoption of the spiked 

 surface. And the reason is obvious : the same benefit the spikes 

 profess to confer is attainable by simpler and more trustworthy iijetbods. 

 Who is there among practical men prepared to maintain that a mo- 

 derate extension of the flue surface, or the employment of an appliance 

 to enable the feed water to extract the superfluous heat from the hot 

 air proceeding to the chimney — expedients whose simplicity is obvious, 

 and whose elKcacy has been established by the experience of a cen- 

 tury — is not more safe, and in every respect greatly preferable to this 

 expensive, inconvenient, and untested innovation ? A flue surface 

 studded with spikes would not only be more difficult to make tight in 

 the first instance, but would be almost impossible to keep tight in 

 those situations where the heat was great. It is well knonu to practical 

 men, that in the furnaces of marine and locomotive boilers even rows 

 of rivets are objectionable, and are as much as possible avoided ; and 

 the head or nut of a bolt in the same situation is found to be very 

 quickly worn away bv the rapid oxidation its high temperature 

 induces. If such be the fate of a piece of iron projecting only an 

 inch from the surface of the plate composing the flue or furnace, what 

 might be expected to become of a piece of iron in the same situation 

 which projected three inches above the plate surface ? How long 



2 E 



