Before the Shar i Society of Engineers and Architects, Mr. W. F. Tyler on Wednesday night read a carefully prepared paper emphasising an aspect of the centigrade system, which is not: always considered when the establishment of scales and tables on a metric basis is under review.
Mr. Tyler's main ent was that the numeral 10, which has been almost universally adopted as the radix number in notations is not, after all, the best adapted for that purpose. Constant familiarity has invested it with some supposed inherent characteristic which does not actually belong to it, and it was to the setting aside of this prejudice in its favour that Mr. Tyler directed his endea-yours. The reason, he said, why ten has been so generally used is the fact that mankind has ten fingers, which have been used from the very earliest time, possibly from the very dawn of articulate speech, for the purpose of indicating numbers. Corroborations of this statement was found in the fact that among the Esquimaux the radix used is four, and the numbers are signified in gesture by the two arms and two legs, the four of which are the origin of the radix. The clumsy mittened hands of the Esquimaux make it impossible to use the fingers as numeral signs in that frost-ridden region. Mr. Tyler went in some detail into a comparison between the decimal and the quaternary notations, and: incidentally established the principle that twice eight is not necessarily one-six (16). In a notation, the index of which was 9, twice eight would be represented by one-seven (17), to which the name "seventeen" might perhaps be given. By many examples Mr. Tyler demonstrated that it is just as easy to deal with numbers in one notation as in another, multiplication, division, and the manipulation of fractions are exactly analogous processes. That conclusion once arrived at, the choice of a unit was next considered and the speaker instituted a comparison between several. The chief requisite, as he pointed out, is that the unit, should be capable of subdivision into those fractions of the whole which are in most general demand. Under this test ten comes out poorly, being divisible only by halves and fifths; eight is divisible by halves and quarters; 16 by halves, quiarters and eizhths; while 12 is divisible by halves, thirds, quarters, and sixths; other and larger numbers are capable of further division, and had man a sufficiently comprehensive mind, 60 might be found a convenient radix, as the smallest number admitting of halves, thirds, quarters and fifths; but under the limitations which have to be accepted, Mr. Tyler made out a strong case for 12.' In setting up a new and universal system of measurements and scales the ideal remedy therefore would be, in his opinion, the adoption of 12 for both the subdivision of scales and for the radix of the notation to be used. That that, however, was practically im-Possible, was one of the annoying conclusions to which the investigator was almost certainly driven. The other alternative were the adoption of the declinal system in spite of the difficulties in the way of subdivision; or to adopt a duodecimal subdivision for all scales, in spite of the difficulty, while continuing to use the decimal notation. The question resolved itself into the choice of the greatest disadvantage, between arithmetical difficulties and imperfect subdivision. As to that, Mr. Tyler held, there -could
Text read from the printed page and cleaned; old type leaves the occasional error.