748 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1908. 



Once when Lord Kelvin was in a chatty mood I asked him point- 

 blank how it occurred that he was not senior wrangler. His blue 

 eyes lighted up as he proceeded to explain that Parkinson had won 

 principally on the exercises of the first two days, which were devoted 

 to text-book work rather than to problems requiring analytical inves- 

 tigation. And then he added, almost ruefully, " I might have made 

 up on the last two days but for my bad generalship. One paper was 

 really a paper that I ought to have walked through, but I did very 

 badly by my bad generalship, and must have got hardly any marks. 

 I spent nearly all the time on one particular problem that interested 

 me, about a spinning top being let fall on to a rigid plane — a very 

 simple problem if I had tackled it in the right way — but I got in- 

 volved and lost time on it and wrote something that was not good, 

 and there was no time left for the other questions. I could have 

 walked over the paper. A very good man Parkinson — I didn't know 

 him personally at the time — who had devoted himself to learning 

 how to answer well in examinations, while I had had, during pre- 

 vious months, my head in some other subjects not much examined 

 upon — theory of heat, flow of heat between isothermal surfaces, 

 dependence of flow on previous state, and all the things I was learn- 

 ing from Fourier." And then he drifted off into a talk of his early 

 papers, and to the mathematical inference (as the result of assigning 

 negative values to the time t) that there must have been a creation. 

 " It was," he continued, " this argument from Fourier that made me 

 think that there must have been a beginning. All mathematical con- 

 tinuity points to a beginning — this is why I stick to atoms * * * 

 and they must have been small — smallness is a necessity of the com- 

 plexity. They may have all been created as they were, complexity 

 and all, as they are now. But we know they have a past. Trace 

 back the past and one comes to a beginning, to a time zero beyond 

 which the values are impossible. It's all in Fourier." 



On leaving Cambridge Thomson went to Paris and worked in the 

 laboratory of Regnault at the College de France. He was here four 

 months. There was no arrangement for systematic instruction, and 

 Thomson's principal occupation was to work the air pump to make a 

 vacuum in one of two large glass globes which Regnault was Aveigh- 

 ing against one another in some determinations of the densities of 

 gases. He made here the acquaintance of Biot and of Sturm and 

 Foucault, of whom he spoke in terms of admiration. Returning, he 

 was awarded a college fellowship of £200 a year. 



Thomson was now 21 years old, but had already established for 

 himself a growing reputation for his mastery of mathematical phys- 

 ics. Ho had published about a dozen original papers, and had 

 gained experience in three universities. In 1846 the chair of natural 



