Measurement of Length 11 



this book are of similar character to those suggested 

 to be done by the student. 



(c) The description of the experiment is best written 

 in the past tense. On no account should the wording 

 of the text-book be slavishly followed, and it is 

 sometimes of service to read through the directions 

 in the text-book and then to perform tlie experiment 

 from memory, writing out an account of it entirely 

 in one's own words. Further reference to the text- 

 book should only be made in order to guard against 

 the omission of any important portion of the work. 



(d) Where a number of readings, or other observa- 

 tions of similar character, are made, they should be 

 tabulated as affording a readier means of comparison 

 of any variations noted. 



(e) In stating numerical results, attention should 

 always be paid to the accuracy with which measure- 

 ments have been made, and no result should be so 

 stated as to suggest that a greater accuracy has been 

 attained than that of the least accurate measurement 

 made. Thus, in the experiment as written out in 

 paragraph 9, the mean result of the measurement of 

 AB in inches is given as 3*74 inches, although the 

 result of dividing the sum of the measurements by 

 3 gives 3*7433 .... Here our measurements were made to 

 the nearest hundredth of an inch, and hence our result 

 must also be stated with that degree of accuracy. To 

 say that the length of AB is 3- 743 inches would mean 

 that it is not less than 3 7425 and less than 3- 7435 

 inches, i.e. that the maximum error of measurement 

 is O'OOOSin. and hence that the whole measurement is 

 correct within -goW m - We cannot by the unaided eye 

 measure length with such accuracy as this. That AB 

 is not less than 3 - 735 inches and less than 3'745 inches 



