By means of portable cesium-beam clocks, has been cor- related to 1 microsecond at many of the world's timekeep- ing centers. A comparison of four of the world's best-known 'long-beam' frequency standards has also been made. Fig. 2 (at left). Dr. Yoshikazu Saburi (cen- ter) of Radio Research Labs, home of Stand- ards Broadcast Station JJY, Japan, discusses time and frequency measurement results with -hp- engineers Richard Baugh and Lee Bodily.

Fig. 3. Operation of trau- eling cesium-beam clock @ checked by -hp- En- gineer Richard Baugh before start of ocean- spanning journey. Cesi- um-Beam Frequency Standard (middle instrument) has self-check- ing feature that enables adjustment of cesium-beam resonator for correct operation without reference to any other frequency standard.

THEFEASIBILITY of transporting com- scales maintained by widely-separated microseconds of the first comp pact ‘atomic’ clocks in continuous timekeeping laboratories have been made eight months ago. This is e operation to distant points by com- correlated within a microsecond, an mercial airliner or other conventional accuracy far higher than that possible transport was established by a flying by high-frequency radio broadcasts of Another accomplishment of clock experiment performed in the time signals which, at intercontinental periment was the verification summer of 1964.l That experiment distances, have only been able to correlated the time of day within a achieve accuracies in time comparisons nization experiment cond microsecond between the Observatoire of about 1 millisecond at best. Indica- of the satellite Relay 11. de NeuchLtel in Switzerland and both tive of the accuracy achieved in the lat- the U. S. Naval Observatory and the est flying-clock experiment, one of the track Radio Tracking Station at the National Bureau of Standards in the portable clocks changed less than Goldstone tracking site in the Califor- United States. It also established 10 pec during the 29 days between nia Mojave Desert and the trackingsta- within 200 microseconds the propaga- two measurements made with respect tion operated for NASA at Kashima, tion time of radio timing signals from to the Master Clock at the Naval Ob- Japan, by the Japanese Radio Re- Beltsville, Md., to NeuchLtel, and it servatory, which establishes Universal search Laboratories. Preliminary re- provided an opportunity for intercom- (Earth rotational) time in the United sults from the U. s. Naval Observator paring laboratory-constructed ‘long- States. The other clock changed less indicate that the exp beam’ cesium frequency standards by than 1 pec in a 23-day period with re- also used VLF tran way of the traveling cesium-beam spect to the NBS UA Time Scale, a standards. time scale based on Atomic time but A new flying clock experiment has offset slightly to approximate Univer- now been performed with the travel- sal time. ing cesium-beam clocks, an experi- The experiment repeated the com- ment that involved more time-keeping parison made last summer between The experiment provided an oppor- centers than any clock correlation ex- time scales in the United States and tunity for frequency comparisons be- periment conducted so far. Standards that maintained by the Observatoire tween the traveling standards and four facilities in Japan, Hawaii, Canada, de NeuchLtel in Switzerland, but in so laboratory-constructed “long-beam” and most of the countries of western doing, it provided further information cesium standards, five commerci Europe were visited in the experiment. by disclosing any relative drift in the built cesium-beam standards, and In all, time and frequency were corre- time scales. As described later, the com- one hydrogen maser. The inte lated at 21 different faciIities in 11 parison shows excellent agreement be- different countries. tween the time scales of the Observa- As a result of the experiment, time toire de Newhiitel and the National - Bureau of Standards in Boulder, Colo- 1 Alan S. Bagley and Leonard S. Cutler, “A New Perform- ance of the Flying Clock Experiment,” Hewlett-Packard rado, the two scales being within 50 Annual Symposium an Frequency Control, 1 Journal, Val. 15, No. 11, July, 1964. Iished. 1

of cesium-beam standards - s that were independently de- and constructed at widelysep- ed points of the globe - provided her evidence that cesium-beam res- ary frequency stand- accuracy. For instance, frequency comparison cesium-beam standard in e traveling clocks and the er cesium-beam standard of the s Horological Research Labora- Laboratoire Suisse de Recherches res) in Neuch2te1, considered ne of the most accurate fre- y standards in the world, agreed 1011 with a compari- etween the same traveling stand- and one of the long-beam cesium Fig. 4. “Long-beam“ cesium resonator at National Research Council, Ottawa, Canada, is typical of high-precision, laboratory-constructed frequency standards dards (NBS 11) at the National intercompared during flying-clock experiment. Cesium-beam standards in both of Standards in Boulder, Colo- traveling clocks agreed with this standard to better than 1 part in 10“. o separate comparisons made apart between the traveling developed compact cesium-beam fre- time. The Model 115BR also generates d and NBS I1 agreed within 1 quency standards, because of their rel- a 25-pec pulse once per for 1012. This performance attests atively small size and modest power re- measurements of the time correspond- stability and reliability of port- quirements, now make it possible to ence between it and another clock. cesium standards in a traveling transport atomically-controlled clocks The Frequency Standard and Clock nment, as well as confirming the by conventional carriers. The clocks were powered by special power sup- ental accuracy of cesium res- used in the experiment described here plies, designed for the clock experi- were each controlled by an -hp- Model ment, which contain rechargeable CESIUM-BEAM CLOCKS 5060A Cesium-Beam Frequency Stand- nickel-cadmium batteries capable of s driven by quartz crystal os- ard, a relatively small, lightweight (65 powering the instruments for several tors have been used in many trav- lbs.) instrument requiring minimum hours. The power supplies can also eling-clock time synchronization ex- power (40W at 26 V dc) but neverthe- power the instruments and recharge periments. Higher accuracy is obtained less capable of high accuracy (&2 x the batteries when operating from ac with clocks controlled by atomic res- 10-l1) and long term stability (&1 x sources ranging from 100 volts to 280 onators, however, because the micro- 10-11).* volts at any frequency from 50 to 400 wave spectral lines characteristic of The traveling clocks each used an cis, or from 6 or 12-V dc sources. The ertain atoms are fundamental physi- * These Standards use a compact cesium-beam -hp- Model 115BR frequency divider tube developed especially for them by the cal constant3 that can be reproduced Quantum Electronics Division (Bomac) of Varian to integrate frequency and display Associates, under contract to Hewlett-Packard. by independent means anywhere on earth without reference to another pri- mary standard. Most of the world’s frequency standards now use cesium- beam resonztors as an absolute refer- ence since a certain cesium spectral line at present provides the most prac- 1 means for controlling frequency by a fundamental atomic constant. 5 10 15 20 25 2 7 I2 I7 FEE (Clock No 4 Reset)> MAR The International Fundamental Unit TIME + 0 Clock No 3 tick minus Clock No 4 tick. of Time (the Atomic Second) is based Phase difference between 100 kcls outputs ot CtockNo. 3 and ClockNo 4 on this particular cesium transition Fig. 5. Time displacement between once- dividers. Discontinuity on Feb. 20 results per-second “ticks” produced by flying from restart of Clock No. 4 (see text). Until recently, however, atomic clocks as measured during trip. Phase dis- Slope of graph during March indicates clocks required the use of cargo planes placement between 100-kcls outputs of that clocks agreed within 2.4 parts in IO”. transportation because of their Cesium Beam Frequency Standards was During February, slope is equivalent to 4 also measured to provide correction factor parts in lo“, well within specified d power requirements. Newly- for any phase shifts in clock frequency accuracy of clocks.

*3* multipled to higher greater measurement peated measurements o the beat note provided

tween the two standards.

INITIAL ADJUSTMENTS

Feb. 2, 1965 at Palo Alto, Califor where the Cesium-Beam Standard

were aligned independently of one other. The cesium-beam “C” field,

Fig. 6. Time kept by traveling clocks is compared to Master Clock at U. S. Naval Observatory, which establishes time in United States, by -hp engineers Baugh and Bodily and by William Markowitz, Director of Time Service Division of U. S. Naval Observatory. Time of once-per-second ticks of traveling clocks, which had been set according to -hp- House Standard three days earlier, differed from time of Naval Observatory ticks by less than 80 psec. cedure that uses the cesi to check on its own perfo instruments may also operate, without ated by the two clocks being compared. Cesium-Beam Frequency recharging the batteries, from 24-V dc Electronic counters were used to meas- sources. Thus it was possible to power ure the time interval between ticks the clocks from the aircraft on which with a resolution of approximately 0.1 required, without reference to an they were flown, from storage batteries psec. other standard. or automobile electrical systems, from Frequency comparisons were made commercial power lines anywhere in in some cases by measuring the phase which derive the 9192-Mcls c the world, or, when necessary, from difference between waveforms pro- resonance frequency from the the internal batteries. duced by the two standards being com- controlled 5-Mc/s oscillator MEASUREMENT TECHNIQUES pared. A plot of many of these meas- were set to introduce a frequen Time scales were correlated by meas- urements over a period of time showed of -150 x 10-10. Thus, the 0.1 uring the time displacement between the relative frequency drift of the two the one-second electrical “ticks” gener- standards. TABLE 1. OVERSEAS FACILITIES VISITED DURING FLYING-CLOCK EXPERIMENT

COMPARISONS PLACE FACILITY TIME FREQUENCY mation of UT2 time, a time sc Koganei, Japan Standard Broadcast Station JJY X X on the rotation of the earth Kashima, Japan RRL Tracking Station X Maui, Hawaii Standard Broadcast Station WWVH X keeping institutions. Herstmonceux Castle, Eng. Royal Greenwich Observatory X The clocks were then set to t Teddington, England National Physical Labs X X WW-referenced time maintained Bagneux, France National Center for X X Communication Studies (CNET) Paris, France Paris Observatory Stockholm, Sweden Swedish Nat‘l Defense Research Institute Copenhagen. Denmark Royal Danish Air Force Calibration Centre with respect to e Hamburg, Germany German Hydrographic Institute (DHI) Braunschweig, Germany National Bureau of simplify time interval Physics and Technology (KB) tween the clocks th Neuchstel, Switzerland Swiss National Observatory NeuchBtel, Switzerland Swiss Horological Research Lab

The Hague, the Netherlands Post Telephone and Telegraph X

Brussels, Belgium Royal Observatory of Belgium X 3 Leon2ird S. Cutler “Exar Lines of a Cesiuk-Beam Ottawa, Canada National Research Council X Synth esizer,” Hewlett-Pa Dec., 1963. (=locksNos. 1 and 2 which were used - in the prcvi~usexperiment.) THE JOURNEY The route of the recent flyingclock esp-ment is diagrammed in Fig. 1 and the overseas facilities visited are summarized in Table I. The clocks were flown on Feb. 5,1965 to the U. S. Naval Observatory in Washington, I). C., for a measurement of the differ- ence between time as kept by the trav- eling clocks and the time kept by the master clock at the Naval Observatory, which defines time in the United States. The cesium-beam standards were also compared in frequency to a hydrogen maser maintained by the Fig. 8. Traveling clocks arrive at Royal Greenwich Observatory, Herst- monceux Castle, England, home of . Naval Research Laboratory. The clocks then traveled to the tracking station in the Mojave Desert From the Mojave Desert, the clocks ing circuitry. Unfortunately, a defect for a time comparison and thence to were taken to Boulder, Colorado, for in the lamp circuit shorted out the Japan. Frequency and time compari- time and frequency comparisons at the circuit and shut down the clock. This sons were made at JJY, the Japanese National Bureau of Standards. During required a restart, including readjust- Standards Broadcast Station, and one transfer, however, Clock No. 4 experi- ment of the “C’ field, after the mal- clock was taken to the Kashima track- enced a transient condition which ac- function was corrected, but, since time ing site for a time comparison. tivated a warning lamp in the monitor- and frequency comparisons were made On returning from Japan, a stop was made in the Hawaiian Islands for 2 time comparison with NBS standard broadcast siation WWVH. The clocks TABLE 11. COMPARISONS BETWEEN TIME SCALES AND were then returned to the Mojave ARBITRARILY SELECTED REFERENCE (NBS UA) tracking station. DATE TlMEtUn FACILITY COMPARISON 2 Feb. 1906 -hp- House Standard -498 psec 12 Feb. 0251 U. S. Naval Observatory -326 psec 17 Feb. 0710 Radio Research Lab +lo01 psec 17Feb. 2340 WWVH -231 psec 21 Feb. 1608 Traveling Clock No. 3 -406.3 psec 21 Feb. 1609 Traveling Clock No. 4 -308.2 PS~C 22 Feb. 1333 U. S. Naval Observatory -329 psec 22Feb. 1743 WWV -483 psec 24 Feb. 1302 Royal Greenwich 0bs.a +5019 psec 25 Feb. 1428 Nat‘l Physical Labs@ +705 pse~ 26 Feb. 1012 Centre Nat‘l d’Et. Tele. -242.198 ms 26 Feb. 1454 Paris Observatory -752 psec 27 Feb. 2246 Swedish Nat‘l DRL -883 psec 2 Mar. 1243 Royal Danish AF Cal. Lab +115.952 ms 3 Mar. 0817 Deutsche Hydro Inst. -398.228 ms 5 Mar. 1027 Standard Broadcast HBN +1353 psec 11 Mar. 1630 Nat‘l Res. Council@ +6 psec 13 Mar. 1513 U. S. Naval Observatory -342 psec 14Mar. 1603 WWV -481 psec 1 Otlret about SWrrec from BSUA to tacilitateme~s~ement~ 16 Mar. 1545 Traveling Clock No. 3 -411.9 psec 2 Maintainedwith respect to Universal lime by astronomicalobsarratims 16 Mar. 1544 Traveling Clock No. 4 -308.9 psec 8@ cntmlled by atomfc tranutias but with offset that approximates Uninrsat he 17 Mar. 0611 -hp- House Standard -495 psec Results. determined by flying clocks, have been adjusted to Fig. 7. Time displacement between clock as account for constant time drift of traveling clocks, as determined determined by traveling clocks. Vertical scale by time closure at NBS, Boulder. Time scale differences, which shows time rektwnship of once-per-second ‘‘ticks‘‘ may be as much as tenths of seconds, have been made large in many cases to facilitate measurements. In all cases, time differ- of indicated time stan-dard with-respect to “ticks” ences have been known for some time but are now known more of NBS UA, which was arbitrarily selected as ref- precisely because of flying-clock measurements. erence time scale. As described in text, -hp House Standard (-hp- HS) is maintained in close agree- ’ Maintained with 5-ms offset with respect to internationallycoordinated time ment with WWV and traveling clocks (-hp 3 and signal broadcasts. 4p4) initially were offset 100 and 200 w respec- Maintained on Atomic time. tively with respect to -hp HS. Clock set according to traveling clocks. Fig. 9. Frequency comparisons between traveling clocks and other cpsiumbeam standards are plotted here as frequency differences in parts per 10". Offsetof -150 x 10" of stand- ards operating on has been removed to Measurements of the time dis show only fundamental fre- quency differences.Results of ment between the one-second ti measurements on all stand- an arbitrarily selected reference ards shown are well within scale and the time scales of va known accuracies of individ- ual standards.

termined from the actual time FREQUENCY COMPARISON (A f/fo Jx,lOW o Comparison to -hp- 3 ference between the flying clock 0 Comparison to -hQ- 4 urement. Assuming a constant between the clocks and NBS standards the clocks were taken to the continent following restart, the shutdown did for time and frequency comparisons not affect subsequent measurements. in the cities and facilities listed in The clocks were then returned to the Table I. made with the reference scale, in thi U. S. Naval Observatory for a time On returning from Europe, the case the NBS UA time scale. closure. The clocks were also com- clocks were taken to Canada for time pared to time as maintained by the and frequency comparisons at the Na- master clock at NBS station WV; tional Research Council in Ottawa. Fig. 9. As was true of the tra The clocks then proceeded to Lon- They then returned to Washington for standards, those standards den0 don for a time comparison at the Royal another time closure with the U. S. Greenwich Observatory and a fre- Naval Observatory and with WWV; spect to UT2 time were arranged quency comparison at the National The clocks then went to Boulder for generate standard frequencies offset Physical Laboratory. From London, yet another time closure and frequency -150 parts in 10'0 with respect to

TABLE 111. COMPARISONS BETWEEN CESIUM-BEAM FREQUENCY STANDARDS AND TRAVELING CLOCKS

Feb. 19 Mojave No. 5* 12.4 cm Commercial 0 x 10 l2

FeB. 1617 RRL (lapan)" 12.4 cm Commercial -3.3 x 10" 29 samples (with -hp- Clock -(150.012 2.018) x 10 Io -(150.018 k.012) xlO'' 33 samples (with -hp- Clock 200secondseach 164 cm Lab type 57 samples (with -hp-Clock -(150.002 f.013) x 10 In -(150.005 2.011) X 10 Io 58 samples (with -hp- Colck 200 seconds each

Feb. 25 HPL (Teddingtonl 277 cm Lab type -(149.81 e.03) X loLa -(149.75 5.03) X 10 Io Rb1-hr. transfer phase standard comparison via Feb. 20 Swedish yat'l Def. Res. Lab. 12.4 cm Commercial +5.1 X 10 Via Rb transfer standard 90 cm Commercial -149.67 X loto 12-hr. phase comparison -(150.09 f .04)x 10" 7-hrs averaging 408 cm Lab type

Mar. 11 Nat'l Res. Council 200 cm Lab type (Canada) time. The diagram of Fig. 9 49 psec. Agreement within 49 TABLE IV. TIME CLOSURES n adjusted to account for the pec of the previous measurement 273 Comparisons derived by time closures per- ce in time scales and reflects days earlier is equivalent to an average formed in the recent flying clock experiment the difference in the cesium-beam are shown in the following. The frequency dif- y frequency difference of 2.1 parts in ference is determined from time difference by itation frequencies. 1012,certainly a remarkable result but the relationship: Af/f, = -At/T. in view of the fact that none of the Traveling Clocks vs. U. S. Naval Observatory Master Clock frequency standards involved are Feb. 22 Clock No. 4 leads a time when im- judged to be this accurate, it can best U. S. N. observatory by 20.4 psec Mar. 13 Clock No. 4 leads be considered a fortuitous circum- U. S. N. Observatory by Spsec stance at the present time. Many more 19 days -13.0 psec -13.0 psec/I9 days = 7.9 x 10“ plete. It is interesting to note, experiments will be required to estab- average frequency difference. ever, that the frequency compari- lish the time correspondence with the Feb. 12 Clock No. 3 leads U. S. N. show all the cesium-beam stand- Observatory by -78.5 psec (A) potential accuracy of the measurement Feb. 22 Clock No. 3 leads U. S. N. to be well within their present technique. Observatory by -77.6 psec (6) Mar. 13 Clock No. 3 leads U. S. N. ified accuracy. Those of recent de- To provide the data necessary for Observatory by -68.7 psec (C) time correlations on a long-term basis, (A) - (6) = -0.9 psec in 10 days = 1.0 X 10 It average frequency difference tinuing improvement in the accuracy Hewlett-Packard plans to repeat the (6) - (C) = -8.9 psec in 19 days = achieved by atomic beam frequency flying clock experiments periodically. 5.4 X 10 ‘*average frequency difference Traveling Clocks vs. CHECK ON VLF PHASE COMPARISONS NBS UA (Boulder) Time Scale* The Flying Clock Experiment also Feb. 21 Clock No. 3 leads COMPARISON OF SWISS OBSERVATORY NBS UA by -406.3 psec AND NBS TIME SCALES confirmed the usefulness of VLF Phase Mar. 16 Clock No. 3 leads NBS UA by --psec As mentioned earlier, time between Comparisons4 for maintaining accu- 23 days 5.6 psec rate time, once that time between sta- 5.6 psec/23 days = -2.8 parts in average frequency difference. tions is definitely established. During 10” d and the National Bureau of Feb. 21 No. 4 leads NBS UA by -308.2 psec rds in Boulder, Colorado, was August of 1964, the -hp- Standards Mar. 16 No. 4 leads NBS UA by --psec ted during this experiment as a Clock was set to agree very closely with 23 days 0.7 psec 0.7 psec/23 days = -3.4 parts in - WVtime by means of another flying lo1%average frequency difference. clock. Since then, phase comparisons Traveling Clocks vs. between the crystal oscillator which WWV (Beltsville) Master Clock t in June, 1964.l The results of the Feb. 22 Clock No. 3 leads WWV by 76.7 psec drives the clock, and the VLF standard Mar. 14 Clock No. 3 leads WWV byarsec frequencies broadcast by NBS stations 20 days 6.6 psec 6.6 psec/20 days = -3.8 parts in WWVL and WWVB, have provided 10” average frequency difference. June. I964 Flying Clock No. 1 leads information for periodic readjustments Feb. 22 Clock No. 4 leads WWV by 174.9 psec NBS UA Time Scale by . . . . 1I50lrsec of the oscillator, thus maintaining the Mar. 14 Clock No. 4 leads WWV by-psec Station HEN’ time ticks lead Clock No. 1 by ...... 440psec clock in agreement with WWV time. 20 days 2.5 psec Neuchhtel Observatory Time Scale 2.5 psec/20 days = -1.4 parts in leads HBN time ticks by . . . . 170pec A measurement at Beltsville, Md., IO” average frequency difference. Neuchatel Observatory Time Scale Traveling Crock vs. -hp- House Standard leads NBS UA by ...... I760psec on Feb. 21,1965 showed that traveling Swiss Standards Broadcast Station Feb. 2 Clock No. 3 leads * Clock No. 3 was 77 microseconds -hp- Standard 96.1 psec The recent experiment provides in- ahead of WWV-referenced time. From Feb. 17 Clock No. 3 leads -hp- Standard Spsec formation for comparing the two time the known time drift of -hp- Clock NO. 42 days 13.2 psec scales over the elapsed time between 3, it was known that No. 3 was 90 mi- 13.2 psec/42 days = -3.6 parts in 10‘‘ average frequency difference. experiments. The results of the 1965 croseconds ahead of the Hewlett-Pack- * Measurement actually made against NBS Working ard Standards Clock on that date. The Standard No 8 then corrected to NBS UA difference between the -hp- Clock and WWV on Feb. 21 is therefore 90-77 or only 13 microseconds. This close ...... 1762psec correlation, which was maintained time, once a time correlation is estab- over a 6-month period since establish- lished...... 17llasec ment of time on the -hp- Standards WORLD TRAVELERS by Clock No 3 Clock No. 4 lags Clock, was achieved by occasional re- The trip covered some 35,000 miles NBS UATime Scale by . . . -309psec . adjustment of the clock oscillator ac- by air and 2,300 miles by car. Nine air- cording to information supplied by lines and eight different kinds of air- VLF phase comparisons. This demon- craft were involved. The clocks usually strates the usefulness of VLF phase traveled %-fare as “children” (230 lbs. comparisons for maintaining accurate each) and occasionally as excess bag-

4 Dexter Hartke “A VLF Comprator for Relating Local gage or cabin freight. Electrical power Frequency to il. S. Standards, Hewlett-Fackard Jouml, Vol. 15, No. 2, Oct., 1964. was usually obtained from the galley,

*7* of their passengers, (A second a however, was more understandi flew the clocks from London to Paris only slightly later than originally British personnel at Her planned.) Otherwise, the reaction of lay per- to sonnel to the instruments in some cases ig was nothing short of spectacular. In into the building and get them operat- London, for instance, the police had ing on ac power once more. been informed that the instruments All in all, the people in the various contained “atomk” material and since laboratories that were visited were the police wished to take no chances, most hospitable and cooperative, will- the voyagers found themselves accom- ing to give up free evenings and Sun- panied by a motorcycIe escort from the days to make measurements if the airport to the first night’s stop. travel schedule so required. Most were Disaster nearly befell the expedition keenly interested in the experiment during the overnight stop at the Eng- being performed and requested a re- lish hotel, however, despite all the turn visit so that long-term variations solicitations of police and others. AI- in time scales may be determined. AS Fig. 10. Dr. L. Essen, who developed though the engineers had brought an mentioned earlier, such return visits worlds first Cesium’s3 atomic-controlled are planned for the future. frequency standard at National Physical assortment of adapters for the variety Laboratories, examines compact cesium- of electrical outlets they expected to ACKNOWLEDGMENTS beam frequency standard that controls encounter, they did not have One that traveling clock. We at Hewlett-packard wish to ex- fit the wall outlets in thevacant confer- ence room where the clocks were to be or from the self-contained batteries if lodged. This required a dismantling of individually - at all the overseas the trip was short. one of the light fixtures to permit at- The inherent ruggedness of the trav- tachment to the power cord. During eling docks was demonstrated by the the night, however, a beIlboy let him- Tracking Station, and facilities trip from Tokyo to the tracking station self into the locked room to clean the National Bureau of Standards, at Kashima, which occurred without ashtrays - and he conscientiously mishap. The 100-mile route was over turned out the lights upon leaving. a tortuous, unimproved road that re- The cIocks continued to operate on with gratitude their encouragement quired five hours to negotiate each their internal batteries, course, but of and valuable assistance, which con- way. The time closures which included these had already been partially dis- tributed immeasurably to the success this part of the trip show that the clock charged during the transfer from the of the experiment. was unaffected by the journey. aircraft and there was barely enough The time taken to travel to the re- charge left the next morning to keep -LaThare N.Bodily mote tracking site, incidentally, re- quired an overnight stay at a non- American-oriented Japanese inn, a rather pleasant, other-worldly experi- ence for the American engineers ac- companying the clocks. The use of the word “atomic” to describe the nature of the clocks had unfortunate repercussions even though the clocks had nothing fissionable within them. Curious fellow passen- gers on the aircraft often changed to seats further away when they learned what those space-age-lookingmachines were. The flight from London to the continent was deIayed because the air- Iine which was to transport the clocks had heard about the devices being g “atomique” and balked at transport- ing devices that might affect the safety

@ p P L: HP Archive

This vintage Hewlett Packard document was preserved and distributed by www. hparch ive.com Please visit us on the web !

Thanks to on-line curator: Kenneth Kuhn for supplying and scanning this vintage document.