106 



Mr. J. S. Macdona]d. Studies in the 



21 amperes through the electromagnets of the cycle, as was known from the 

 results of calibration experiments most kindly performed for me in the 

 Electrical Engineering Department of the University by Mr. Bissett, under 

 the supervision of Mr. Crapper, to both of whom I am greatly indebted. 



When a considerable number of such experiments had been performed 

 with results all placed within certain limits of the scale of possible heat- 

 production I was suddenly introduced to two simultaneous phenomena, 

 (a) the ammeter connected with the circuit of the cycle brake, although 

 mainly preserving an appearance of carrying the required currents, yet 

 showed sudden intermissions during which the current evidently fell in 

 value for short periods of time that seemed to be associated with some 

 regular phase in the revolution of the cycle, and (b) that my experimental 

 results now began to occupy new regions in the scale of possible heat- 

 production. An examination of the electromagnets of the cycle-brake 

 revealed a wire insecurely held within a soldered joint in such a fashion 

 that the total resistance of the joint must have been very variable. 



This fact is my quite sufficient excuse for omitting this particular group 

 of experimental results. It also involved the complete rewinding of the 

 electromagnet, since the fractured wire was buried by other coils, and there- 

 fore very inaccessible. I was thus involved in further consequences since 

 I was obliged to recalibrate the cycle-brake within the laboratory. Two 

 sets of calibration experiments were performed, (1) with 1 ampere, and 

 (2) with 2 - l amperes in the altered brake. I now, however, found myself in 

 each of these cases possessed of a more powerful brake than before, and 

 rather than spend further time in discovering the particular values of 

 current which would give precisely the same brake-power as formerly 

 I proceeded to perform two new groups of experiments, in one group 

 19 kalories being the measure of the mechanical work performed per hour, 

 and in the other case 56 kalories. I have therefore now to deal with four 

 groups of experiments, and from the results it might perhaps appear that 

 this is an advantage rather than a misfortune. 



Before detailing these results I should make some mention of the fact that 

 in addition to omitting a certain group of experiments previously referred 

 to, there are some others which I have deliberately omitted from my list. In 

 the first place, one of my subjects, a laboratory boy, Armstrong, promised to 

 be of great advantage because of bis smaller weight of 44 kgrm., thus 

 occupying a different region of the scale of weights than most of my other 

 subjects. He had, however, no experience of cycling, and was thus, in my 

 opinion, most handicapped in the experiments with only a slight resistance in 

 the brake. Given plenty of work to do on such a treadmill the best way of 



