52 Bulletin Santa Barbara Society of Natural History. Vol. I 



increase in temperature. This again divided by 48, 

 the number of inches in the rod, gave .0000460, or less 

 than 1-20000 part of the length as the increase in terms 

 of the length. 



Many like observations were made at various times, 

 both with these rods and with others, as well as with 

 glass tubes and lead pipes, and generally' with a very 

 fair agreement in the results. 



It was very gratifying to be able thus to ascertain to 

 a certainty that wood, like the metals did expand by 

 heat. But to me the most interesting part of the ex- 

 periment was to see the wonderfully minute space, we 

 were able to measure by such simple means. 



By taking the mean of the different observations, I 

 obtained for the expansion of Spanish cedar in terms of 

 its length, for 30 increase in temperature .0000688, 

 and for that of white pine .0000464. 



The actual extension of each rod, expressed approx- 

 imately in vulgar fractions, was 1-300 and 1-450 of an 

 inch, and these were about the largest spaces measur- 

 ed. What the smallest were it is difficult to say; but 

 the greatest variation of the pine from the mean was 

 about 1-4000 of an inch. If this is the extreme, it is 

 quite safe to say that the probable error in the final re 

 suit, so far as the measurement is concerned, could 

 hardly be one-fourth as much, or less than 1-16000 of 

 an inch — a space that it would require a good micro- 

 scope to see. 



