346 



NATUREi 



[February 13, 1902 



and that a professor of natural philosophy would spend a 

 long time over the inquiry and return an unpractical 

 answer. The problem attacked by the cowl committee 

 happened to be the exceptional tenth case in which, not 

 academic students, but practical men, spent twenty years, 

 a pericrd, indeed, terminated only by the deaths of the 

 members of the committee, without arriving at practical 

 results of the final character looked for. 



As a matter of fact, the flow of air along a three-inch 

 or six-inch tube surmounted by a cowl is a very compli- 

 cated result ; it is no more and no less the measure of 

 the efficiency of the cowl on the top of the pipe than the 

 current through a galvanometer is a measure of the 

 efficiency of a battery cell in circuit with it. If we 

 picture to ourselves a committee endeavouring to pro- 

 nounce upon the relative merits of the battery cells of 

 many inventors by tabulating the deflections which they 

 produce in a galvanometer before (j. S. Ohm had been 

 led, by purely scientific researches, to the law which has 

 been the guide in all such questions since his time, we 

 get an exact analogy of the action of the cowl committee, 

 liefore the efficiency of the best cowl can be effectively 

 represented by a number, it is desirable to settle what 

 purpose the instrument is intended to serve. The pro- 

 duction of a flow of air in a particular combination of tubes 

 is doubtless one purpose, but there are others. Some 

 cowls are intended to keep the rain out of a shaft ; some 

 are picturesque terminations of- flues, as the Italian 

 examples cited by Mr. .'\ckermann in the cowl report ; 

 all, presumably, are intended to bring profit to their 

 makers. As regards efficiency, they might be classified 

 according to their performance with regard to one or 

 other of these widely different purposes, but the classifi- 

 cation would not be strictly scientific. 



To the student of theoretical science they can only be 

 regarded as examples of" apparatus for diverting the 

 kinetic energy of the passing air to produce a flow along 

 the pipe which the cowl surmounts, whether the flow be 

 up or down ; and the laws of transformation of the energy 

 will most effectively describe the behaviour of the cowls 

 from the scientific point of view. The flow along the 

 pipe implies a certain expenditure of energy which must 

 ultimately be derived from the passing air (in the absence 

 of temperature difference), and the primary effect of the 

 cowl may be estimated by the amount of energy which it 

 takes from the wind and diverts to producing or main- 

 taining a flow. We may call this the aeromotive force 

 on the analogy of an electromotive force of a battery 

 maintaining an electric flow m a circuit. 



This conversion of the energy of wind into aeromotive 

 force is in itself a very interesting subject. There is no 

 doubt that by suitable mechanism of the windmill type 

 (hinted at by the .'\rchimedean screws of some revolving 

 cowls) the wind could be made to lift through a chimney, 

 not only air, but also, if desired, the cinders or the coals, 

 and even the firegrate and the hearthstone itself ; but the 

 limits of effectiveness of purely pneumatic as distin- 

 guished from mechanical arrangements would be a very 

 useful and productive subject for study. To take the 

 matter a stage further in detail and ask whether it is 

 possible by any combination of plane or curved surfaces 

 to make the velocity of the air passing over the mouth of 

 a tube greater than the original velocity of the wind, 

 is to suggest an inquiry with important bearings 

 upon many scientific subjects. For example, when 

 wind blows through a truncated cone along the axis 

 of the cone, is the velocity of emergence greater 

 than the velocity of the unimpeded air ? The cowl 

 results do not answer this question, but indicate some 

 suggestions. They show that a large cowl is more 

 effective than a smaller one of similar pattern upon the 

 same pipe ; that the output of a pipe can be considerably 

 increased by surmounting it with an extending cone-piece 

 and adding at the widened end -a " louvre" band attached 



NO. 1685, VOL. 65] 



by a regulated number of " feathers," and above that a 

 cap at a certain distance. Whether this arrangement 

 actually increases the velocity of the air passing over the 

 mbuth of a tube or merely prevents diminution only a 

 master of theory can say. There is no experimental 

 answer to this question. 



Just as the current in a galvanometer gives no final 

 indication of the electromotive force of a battery in its 

 circuit, sothe air current in the tube gives no final indi- 

 cation of the aeromotive force of the cowl. The resist- 

 ances of all parts of the circuit and any accessory aero- 

 motive forces must likewise be accounted for. 



The resistances to be accounted for must refer to all 

 parts of the complete circuit and are more complicated 

 than electrical resistance, for they depend partly upon 

 friction in which the loss of energy is proportional to the 

 first power of the velocity and partly upon the turbulent 

 motion in which the loss of energy is proportional to the 

 square of the velocity. There are also in the circuit of 

 flow arranged in the cowl experiments other aeromotive 

 forces than that due to the cowl. In order to make this 

 clear, it is sufficient to point out that the circuit consisted 

 of the cowl, a length of three-inch pipe, an air meter, a 

 short pipe opening into a box with a silk gauze base, and 

 some opening, either door, or window, or chinks, between 

 the interior of the hut where the observer was accommo- 

 dated and the outside air. If the cowl had been removed 

 the flow would not have ceased ; even if the long tube 

 were removed altogether there would be some flow through 

 the opening, and if the opening were in the side of the 

 hut instead of the top there would still be flow, thus 

 indicating aeromotive forces quite apart from that due to 

 air passing over the cowl or the top of the open tube. 



The cowl committee made no inquiry into the different 

 elements which go to make up the composite effect read 

 on the air meter, and consequently made no attempt to 

 analyse the effect of a cowl into the production of aero- 

 motive force and resistance, just as the effect of a battery 

 may be analysed into the production of electromotive 

 force and electrical resistance. As downdraught is 

 reversed flow, the same kind of uncertainty attaches to 

 the results of the committee's experiments upon this 

 branch of the subject. 



The most confusing results are those for cowls iised 

 as injectors. In the case of exhaust cowls, comparison 

 was made between the flow in three parallel pipes, 

 the middle one carrying the cowl, the other two bare. 

 The direction of the flow was the same in all three and 

 all were fed from the interior of the same hut. When 

 the effect of an injector was to be measured it was 

 mounted upon the middle pipe, the outside pipes still 

 remaining free. The aeromotive force in the injector 

 pipe was, of course, reversed ; one pipe supplied air to 

 the hut and two extracted air from it. In that case flow 

 could go on if the hut were otherwise hermetically sealed, 

 and allowing for the hut being leaky, as it was, inten- 

 tionally so, the network of currents is so complicated that 

 it is difficult to attribute any precise meaning to the re- 

 lation of the flow down the middle pipe and to that up 

 the outside pipes, the relation selected by the com- 

 mittee to represent the effectiveness of the injector. 



I have given some justification of the statement that 

 the results obtained by the committee are no more in- 

 dicative of the characteristic action of cowls than gal- 

 vanometer readings are of the characteristics of batteries; 

 I should like to add some words in explanation of the 

 parallel statement that they are no less so. 



If the resistances in an electric circuit are all main- 

 tained the same and the only changes introduced are 

 successive slight modifications of the battery itself, it is 

 quite possible to obtain from the corresponding galvano- 

 meter readings sufficiently definite information about the 

 efiefcfofthechanges in the battery. Examples analogous to 

 this are afforded in the report by numerous comparisons of 



