June 20, 19 18] 



NATURE 



307 



both the visual and photographic observations. 

 The Cambridge photographs of June 13 are par- 

 ticularly valuable in having oCygni as a corn- 

 parison spectrum, thereby confirming Sir 

 Norman Lockyer's conclusion tha-t many of the 

 enhanced lines which are so prominent in this star 

 are reproduced as bright lines in the spectra of 

 novae. Later observations communicated to us 

 by Prof. Fowler show that this stage continued up 

 to June 16, when his last observations were made. 

 The bright C and F lines of hydrogen, and the 

 enhanced lines of iron A A 517, 502, and 492, have 

 remained the most conspicuous features of the 

 visible spectrum since bright lines appeared, and 

 the diminishing luminosity of the star has been 

 accompanied by a marked reduction in the in- 

 tensitv of the continuous spectrum. 



On June 12 it was found that the bright fringe 

 on the red side of the dark band about A 589 had 

 become a definite bright band, having a narrow 

 dark line near its red edge, the whole group being 

 probably identical with that shown in photographs 

 of Nova Persei taken at the Yerkes and Lick 

 Observatories, which clearly proved that the 

 narrow line was sodium D. Revised estimates of 

 the two bright lines between C and D suggested 

 identity w^ith lines about A 615 and A 62 5 shown in 

 the Yerkes photographs of Nova Persei, and a 

 fainter line was noted about A 641. Two vague 

 brightenings were also observed between F and G. 

 On June 13 the dark band about D was much 

 reduced in intensity, while that about A 560 had 

 become considerably stronger. On June 15 and 16 

 the principal change was the reduced intensity of 

 the continuous spectrum and the consequent 

 greater clearness of the bright lines between 517 

 and D. If the nova progresses at the same rate 

 as Nova Persei, the third stage may possibly be 

 entered upon about the end of the first week in 

 July. 



With reference to the probable distance of the 

 nova, direct determinations of the parallax will 

 necessarily occupy a considerable time. The cir- 

 cumstance that nearly all novae have occurred 

 in the Milky Way, however, furnishes strong 

 evidence that these objects are actually situated in 

 the Milky Way, and, therefore, at distances of 

 the order of, say, 3000 light-years. 



THE NEW SYSTEM OF TIME-KEEPING 

 AT SEA. 



AN article in Nature for April 25 described the 

 new system of time-keeping at sea which 

 was adopted last year by the British,- French, and 

 Italian Admiralties. The Board of Trade has now 

 published a memorandum on the subject, with a 

 coloured chart, adapted from a similar one pre- 

 pared by the " Service hydrographique de la 

 marine francaise." Reference may also be made 

 to useful explanatory articles by M. J. Renaud in 

 "Annuaire du Bureau des Longitudes, igi8, " and 

 in L'Astronomie for April, 1918. 



The chart illustrates the international time 

 system both by land and sea. The countries and 

 NO. 2538, VOL. lOl] 



States that ha\ e adopted it are coloured either red 

 or blue, red denoting Greenwich time, or time 

 differing from it by an even number of hours, 

 while blue denotes a difference of an odd number 

 of hours. A few countries (India, South Aus- 

 tralia, \'enezuela) use time differing from Green- 

 wich time by an odd number of half-hours ; these 

 are coloured violet on the chart. The remaining 

 countries are coloured yellow. Inspection of the 

 chart shows what great progress the international 

 time system has already made. The whole of 

 Europe except Russia, almost the whole of Africa, 

 Japan, Australia, North America, Peru, and 

 Brazil have adopted it. It is not improbable that 

 its adoption by ships may lead to still further 

 extension of it on^ land. 



It is much to be regretted that the Board of 

 Trade uses the term "zones " to denote the regions 

 that keep the same time. The use of this term 

 ought to be confined to the designation of belts 

 parallel to the equator, not at right angles to it. 

 It is much better to employ the term "fuseau," 

 which the French have adopted, unless a suitable 

 English name can be devised. 



Inspection of the chart shows that the boun- 

 daries of the "fuseaux" on land do not strictly 

 follow the theoretical meridians; they frequently 

 deviate to some political boundary not far away, 

 in order to keep the same time throughout a 

 country or State. It is obviously convenient for 

 a ship while in territorial waters to keep the time 

 observed on the adjacent coast ; local tide tables, 

 etc., would be given in this time. But when on 

 the high seas it should change its clocks at the 

 nearest convenient moment to the time of entering 

 the new "fuseau" (say, at the nearest change of 

 watch). 



It is important to note that the change of system 

 is wholly in the direction of greater simplicity. 

 Hitherto there have been two entirely distinct sets 

 of timepieces on board : the chronometers, used 

 in navigating the ship, which keep Greenwich 

 time; and the ship's clocks, used for the ordinary 

 purposes of daily life ; these have usually kept local 

 apparent time, being set about noon on each day, 

 sometimes twice a day if the ship was travelling 

 very rapidly. For the future all clocks on ships in 

 all parts of the world ought to show the same 

 •minute as the gate-clock at Greenwich, the differ- 

 ence being in the hours only. 



The French and English have adopted different 

 modes of numbering the "fuseaux." Both agree 

 to call the Greenwich "fuseau" (extending from 

 7|° W. to 7|° E.) zero; the French number the 

 "fuseaux" to the east of this +1, +2, +3, in 

 succession, up to +23 for the "fuseau " just west 

 of the zero one. These numbers give the correc- 

 tion to apply to Greenwich time to obtain ship's 

 time. The English system uses two series of 

 numbers, each from i to 12, negative to the east, 

 and positive to the west, thus giving the correction 

 to apply to ship's time to obtain Greenwich time. 

 It matters little tvhich is done, provided the 

 system is understood. It is recommended that 

 the " fuseriu "-numlirr he ,il\\ :ivs (lisii!,'i\cd on the 



