;82 



NA TURE 



\AiLgust 15, 1889 



avoirdup fis pounds can be compared with an error not exceeding 

 0'0002 of a grain, and two kilogrammes with an error not 

 exceeding 0'02 of a milligramme. 



The mean solar day is the natural unit of time for the human 

 race, and it is universally adopted among all civilized nations. 

 Our ultimate standard of time is therefore the rotation of the 

 earth upon its axis, and from that rotation we determine the 

 ■errors of our clocks and watches by astronomical observations. 

 For many purposes it suffices to make these observations upon 

 the sun, but when the utmost precision is desired it is better to 

 make them on the stars. Until the close of the seventeenth 

 ■century, quadrants were employed for that purpose, and so late 

 as 1680, Flamsteed, the first English Astronomer- Royal, thought 

 himself fortunate when he succeeded in constructing one which 

 enabled him to be sure of his observed times within thiee 

 •seconds.^ About 1690, Roemer invented the transit instrument, 

 which soon superseded the quadrant, and still remains the be it 

 appliance for determining time. Most of his observations were 

 destroyed by a fire in 1728, but the few which have come down 

 to us show that as early as 1706 he determined time with an 

 accuracy which has not yet been very greatly surpassed. 

 Probably the corrections found in the least square adjustment of 

 •extensive systems of longitude determinations afford the best 

 •criterion for estimating the accuracy of first-class modern time 

 observations, and from them it appears that the error of such 

 observations may rise as high as ± 0'05 of a second. 



During the intervals between successive observations of the 

 heavenly bodies we necessarily depend upon clocks and chrono- 

 meters for our knowledge of the time, and very erroneous ideas 

 are frequently entertained respecting the accuracy of their running. 

 The subject is one upon which it is difficult to obtain exact 

 information, but there are few time-pieces which will run for a 

 week without varying more than three-quarters of a second from 

 their predicted error. As the number of seconds in a week is 

 604,800, this amounts to saying that the best time-pieces can be 

 trusted to measure a week within one part in 756,000. Never- 

 theless, clocks and chronometers are but adjuncts to our chief 

 time-piece, which is the earth itself, and upon the constancy of 

 its rotation depends the preservation of our present unit of time. 

 Early in this century Laplace and Poisson were believed to have 

 proved that the length of the sidereal day had not changed by so 

 much as the looth jiart of a second during the last 2500 years, 

 but later investigations show that they were mistaken, and, so 

 far as we can now see, the friction produced by the tides in the 

 ocean must be steadily reducing the velocity with which the 

 earth rotates about its axis. The change is too slow to become 

 sensible within the lifetime of a human being, but its ultimate 

 consequences will be most momentous. 



Ages ago it was remarked that all things run in cycles, and 

 there is enough truth in the saying to make it as applicable now 

 as on the day it was uttered. The Babylonian or Chaldean 

 ■system of weights and measures seems to be the original from 

 which the Egyptian system was derived, and is probably the 

 most ancient of which we have any knowledge. Its unit of 

 length was the cubit, of which there were two varieties — the 

 natural and the royal. The foot was two-thirds of the natural 

 cubit. Respecting the earliest Chaldean and Egyptian system 

 of weights, no very satisfactory information exists, but the best 

 authorities agree that the weight of water contained in the 

 measure of a cubic foot constituted the talent, or larger unit of 

 weight, and that the sixtieth or fiftieth parts of the talent con- 

 stituted, respectively, the Chaldean and Egyptian values of the 

 mina, or lesser unit of commercial weight. Doubtless these 

 weights varied considerably at different times and places, just as 

 the modern pound has varied, but the relations stated are 

 believed to have been the original ones. The ancient Chaldeans 

 used not only the decimal system of notation, which is evidently 

 the primitive one, but also a duodecimal system (as shown by 

 the division of the year into twelve months, the equinoctial day 

 and night each into twelve hours, the zodiac into twelve signs, 

 &c.), and a sexagesimal system (by which the. hour was divided 

 into sixty minutes, the signs of the zodiac into thirty parts or 

 degrees, and the circle into 360 degrees, with further sexagesimal 

 subdivisions). The duodecimal and sexagesimal system-; seem to 

 have originated with the Chaldean astronomers, who, for some 

 reason which is not now evident, preferred them to the decimal 

 system, and by the weight of their scientific authority impressed 

 them upon their system of weights and measures. Now observe 



' Account of the Rev. John Flamstesd. By Francis Bally. Pp. 45-9 

 •iLonJon, 1835, 4to.) 



how closely the scientific thought of to-day repeats the scientific 

 thought of four thousand years ago. These old Chaldeans took 

 from the human body what they regarded as a suitable unit of 

 length, and for their unit of mass they adopted a cube of water 

 bearing simple relations to their unit of length. Four thovisand 

 years later, when these simple relations had been forgotten and 

 impaired, some of the most eminent men of science of the last 

 century again undertook the task of constructing a system of 

 weights and measures. With them the duodecimal and sexa- 

 gesimal systems were out of favour, while the decimal system 

 was highly fashionable, and for that reason they subdivided their 

 units decimally, instead of duodecimally, sexagesimally, or by 

 powers of two ; but they reverted to the old Chaldean device for 

 obtaining simple relations between their units of length and mass, 

 and to that fact alone the French metric system owes its survival. 

 Everyone now knows that the metre is not the 10,000,000th 

 part of a quadrant of the earth's meridian ; and in mathematical 

 physics, where the numbers are all so complicated that they can 

 only be dealt with by the aid of logarithms, and the constant •tt, 

 an utterly irrational quantity, crops up in almost every integral, 

 mere decimal subdivision of the units counts for very little. But 

 in some departments of science, as, for example, chemistry, a 

 simple relation between the unit of length (which determines 

 volume), the unit of mass, and the unit of specific gravity, is of 

 prime importance ; and wherever that is the case the metric 

 system will be used. To engineers such relations are of small 

 moment, and consequently among English-speaking engineers 

 the metric system is making no progress, while, on the other 

 hand, the chemists have eagerly adopted it. As the English yard 

 and pound are the direct descendants of the Chaldean-Babylonian 

 natural cubit and mina, it is not surprising that the yard should be 

 only 0*48 of an inch shorter than the double cubit, and the avoir- 

 dupois pound only 665 grains lighter than the Babylonian com- 

 mercial mina ; but, considering the origin of the metric system, 

 it is rather curious that the metre is only I '97 inches shorter than 

 the Chaldean double royal cubit, and the kilogramme only 102 

 grains heavier than the Babylonian royal mina. Thus, without 

 much exaggeration, we may regard the present English and 

 French fundamental units of length and mass as representing 

 respectively the commercial and royal units of length and mass 

 of the Chaldeans of 4000 years ago. 



Science tells us that the energy of the solar system is being 

 slowly dissipated in the form of radiant heat ; that ultimately 

 the sun will grow dim ; life will die out on the planets ; one by 

 one they will tumble into the expiring sun ; and at last darkness 

 and the bitter cold of the absolute zero will reign over all. In 

 that far-distant future imagine some wandering human spirit to 

 have penetrated to a part of space immeasurably beyond the 

 range of our most powerful telescopes, and there, upon an orb 

 where the mechanical arts flonrish as they do here, let him be 

 asked to reproduce the standards of length, mass, and time, with 

 which we are now familiar. In the presence of such a demand 

 the science of the seventeenth and eighteenth centuries would be 

 powerless. The spin of the earth which measures our days and 

 nights would be irretrievably gone ; our yards, our metres, our 

 pounds, our kilo^jrammes would have tumbled with the earth 

 into the ruins of the sun, and become part of the deb7-is of the 

 solar system. Could they be recovered from the dead past and 

 live again ? The science of all previous ages mournfully answers. 

 No ; but with the science of the nineteenth century it is other- 

 wise, The spectroscope has taught us that throughout the 

 visible universe the constitution of matter is the same. Every- 

 where the rhythmic motions of the atoms are absolutely identical, 

 and to them, and the light which they emit, our wandering spirit 

 would turn for the recovery of the long-lost standards. By means 

 of a diffi-action grating and an accurate goniometer he could 

 recover the yard from the wave-length of sodium light with an 

 error not exceeding one or two thousandths of an inch. Water 

 is everywhere, and with his newly recovered yard he could 

 measure a cubic foot of it, and thus recover the standard of 

 mass which we call a pound. The recovery of our standard of 

 time would be more difficult ; but even that could be accom- 

 plished with an error not exceeding half a minute in a day. One 

 way would be to perform Michelson's modification of Foucault's 

 experiment for determining the velocity of light. Another way 

 would be to make a Siemens's mercury unit of electrical resist- 

 ance, and then, either by the British Association method, or by 

 Lord Rayleigh's modification of Lorenz's method, find the 

 velocity which measures its resistance in absolute units. Still 

 another way would be to find the ratio of the electro-static and 

 electro-magnetic units of electricity. Thus all the units now 



