168 Mr. F. Jenkin on the Question of the 



in no way affects this part of the question, also that an error in 

 the coefficient of correction for the German-silver copies does not 

 affect the argument ; but I do maintain that when Dr. Siemens 

 and Dr. Matthiessen obtain discrepant values after taking all 

 possible care, we must be allowed to suspend our judgment as 

 to which of the two have obtained the value which is most near 

 to truth. According to Dr. Siemens' s paper, making a correc- 

 tion in his 1864 standards, as he desires, of 0-287 per cent., the 

 B.A. unit = 1*0486 millimetre of mercury of 1 millimetre sec- 

 tion at 0° C. According to Dr. Matthiessen the value is 1*0396. 

 These values are too different to allow us yet to use mercury as a 

 certain means of reproduction. 



As soon as one or more independent observers shall obtain 

 really concordant results, they will be accepted by all as a means 

 of checking the permanency of the material standards already 

 made. Till then, I believe the best check will be found in the 

 comparison of the standards themselves, which can be measured 

 with an accuracy approximating at least to within 0*01 per cent., 

 and are correct within this amount, although, owing to a misun- 

 derstanding and possible ambiguity of the language employed, 

 Dr. Siemens believed they were not even intended to be so. 



I shall not enter into the controversy between Dr. Matthiessen 

 on the one hand, and Mr. Sabine and Dr. Siemens on the other*. 

 Dr. Matthiessen can take good care of himself, and his reputa- 

 tion is too high to allow us simply to accept Dr. Siemens' s results, 

 differing as they do from those of the only other observer who 

 has made researches with similar objects in a distinct laboratory. 



Dr. Matthiessen places greater reliance on reproductions by 



* I should, however, be glad to have an explanation upon one point. 

 Dr. Siemens gives a Table showing that the determinations made with two 

 tubes in 1859, and four tubes in 1860 and 1863, differed less than 0*1 per 

 cent., and the most concordant results were obtained in the latest mea- 

 surements. I venture to point out that, as I understand this statement, 

 it does not prove that the determinations in 1859, 1860, and 1863 were 

 concordant and gave the same unit. It shows that the relative mea- 

 surements of the several tubes on the three occasions were correctly made 

 within 0*1 per cent., and the relative resistances agreed with these measure- 

 ments. But this result would have been obtained, however different the 

 mercury might have been on the three occasions, provided of course the 

 three tubes during any comparison were filled from one and the same source. 

 But if by any chance impure mercury was used in 1859 and pure mercury 

 in 1860, the unit as determined by the apparently concordant observations 

 would be very different, and the difference could only be detected by com- 

 parison with a material standard assumed as permanent. I should be glad 

 to know whether this comparison was made or not ; and if made, with what 

 result. Unless some such comparison has been made, we have no guarantee 

 whatever that the results of the three determinations were really concord- 

 ant. I do not find any mention of this comparison in Mr. Sabine's or Dr. 

 Siemens's papers ; but I may have overlooked it. 



