Atlantic High

by William F. Buckley, Jr.

  Worth a paragraph’s discussion is the last of its listed services, namely Sunrise/Sunset, Meridian Passage. The promise is obvious, namely that you can ascertain the exact time of sunrise and sunset anywhere in the world anytime before the mid-twenty-first century by merely vouchsafing the time, date, and estimated position. Now, this is a convenience that goes potentially beyond the point of telling you when, roughly, you will be needing to put on your running lights. If you can clock the exact moment when the sun disappears below the horizon you can deduce your longitude, even as clocking the exact moment when the sun passes over the meridian gives you your latitude.

  At this point I interrupt to describe a glorious feature of the Navicomp, namely its functional simplicity. In order to work with the star Vega, for example, it isn’t necessary to find, from an inventory of capsules (Texas Instruments) or 1½-inch-long shavings of flexible plastic (HP-97), the item containing Vega, and insert it into the computer. You merely look at a little plastic wallet-size index card under “Vega” and note its code number (49). You depress P, which signals the computer that you are about to orient it toward a particular system, then you depress 49, then you depress D, which is the activator.

  The Navicomp, at the left-hand corner of the display bar, will give you a number in response to any rational activity on the keyboard. Depress P, for instance, and the numeral I appears. Whenever you see I, the computer is saying: “What program do you desire?”

  You depress 207, which is your Sunrise/Sunset, Meridian Passage, and push D to activate it.

  The number 2 is displayed. Whenever you see 2, there is only one datum the Navicomp is asking for, namely: “What is your Assumed Latitude?”

  If you are at Stamford, Connecticut, you depress 4103 (the little index card also informs you of the relevant protocols in the use of decimal points. None between degrees and minutes. But if you wish finer calculations, you may have them—e.g., 4103.2).

  Depress D to enter the datum, and the number 3 flashes. That means: “Give me your Assumed Longitude.” You comply (again, from Stamford): —7332. That minus, by the way, is a chauvinistic giveaway. In fact, an American firm (Litton Industries) owns Plath. But it remains a thoroughly German enterprise, and the scientists who designed the Navicomp believe that normal people reside in the eastern hemisphere (Hamburg: Latitude 53°oo’N. Longitude 10°00’E). So that whereas U.S. navigational computers require you to put a minus sign before introducing east longitude, the reverse is the case with the Plath.

  Depress D. The number 4 appears. It means: “Tell me what time it is in Greenwich.” The theoretical reflexes balk. We are working with a chronometer. Let us imagine that we are making our calculations on May 1, 1979, at 0700 Eastern Daylight Time. When the Number 4 is displayed, the operator has two alternatives. Either he can write in a time which is hypothetical (let us say you desire the time of sunset at Stamford on Christmas Day) or he can log the time at the moment he is doing his reckonings—in this example, as I say, 0700, May 1. He need not bother, in the latter case, to punch out the time—it is always ticking away in the chronometer. So he merely depresses D. The chronometer registers GMT 11-00-00 (add four hours for zone time). The numeral 5 now appears, always the same thing: “Give me the month, the day and the year.” That too is ticking away, prerecorded, so you merely depress D again.

  The Navicomp then sets out to calculate. In about ten seconds it displays on the left the numeral 7, which is a so-called Display Numeral. It isn’t asking for anything now, it is telling you something. You get alternative displays (at intervals of two seconds) of 452.7 and 1849.8. That means that sunrise will occur (add one hour for daylight time, 1/10 of a minute is six seconds) at 5:52:42. And sunset at 7:49:48. (If you then depress D again, you will be given the time when the sun will pass over your meridian—namely, 11:51:18 or, EDT, 12:51:18.)

  You are entitled to wonder: Why am I being asked to give the machine the time at the moment I happen to be inquisitive about the time of sunset? Haven’t I already given the machine the Assumed Position? And the date? Why is it relevant to ask me for the time?

  I inquired rather deeply into the question, and ended with a cable from Hamburg (and eighteen conversations with Plath). I should interrupt to say that if you know the exact time when the sun disappears at an assumed position, and you clock the time when it actually disappears, the difference will give you your longitude. Hamburg cabled, “As the only time-dependent variable in the formula for sunrise and sunset is the declination of sun, which varies only very slowly throughout the day, it is formally sufficient to key in an approximate GMT. For higher accuracy, it is recommended to repeat the calculations for both sunrise and sunset, now using the times achieved during the first calculations (converted from zone time to GMT).”

  We all know that, on May 1, the sun’s declination is gradually increasing. The rate is about nine minutes (nine miles) per twelve hours. So that if you ask the computer for sunset information at seven in the morning, but sunset is at approximately seven in the evening, will your calculations be approximately nine miles off?

  We can check by running the program with 7 P.M. input, as opposed to 7 A.M.

  The answer: Sunrise at 5:52:06. Sunset at 6:50:24. (Compare with sunrise 5:52:42 and sunset 6:49:48.)

  The difference in seconds for sunrise is 36. Four seconds of time equals one nautical mile of distance. If you take chronometer time at 7 A.M., you will miscalculate your longitude by eight miles.

  Let us therefore try out Hamburg’s instructions, and instead of using 7 A.M. time, use instead the sunset estimate yielded by the 7 P.M. time, namely 1849.8. This translates (add four hours) into GMT 22:49:48. What do we come out with?

  The results are identical to the hypothetical 7 P.M. sunset: i.e., 6:50:24.

  Let us now go through the agony of calculating the time of sunset for May 1 at the latitude and longitude of Stamford, Connecticut, through the nautical almanac. I get 18:48:08. For an error of 1 minute 40 seconds. An error the equivalent of 25 miles. Unless I missed something, longitudinal navigation by sunset is not possible with the Navicomp. The time of sunset is close enough only to alert you to the approximate time when to begin your star sights. The question, on which others will wish to meditate, is rendered more mysterious by the testimony of Mr. Leroy Dogett of the National Observatory. He advises that the variable effects of barometric pressure on refraction are such that the almanac rounds off sunsets to the nearest minute. Curses! Anyway, you try it.

  This sunset-sunrise business is the single failing in the Navicomp (other than the unfortunate neglect of Jupiter and Saturn). Correction—there is a second, intensely annoying feature, which is that notwithstanding the highly touted Polaroid case, which protects the instrument and is designed to permit you to read the red figures simultaneously, in fact, you can’t do it. The sun’s brightness completely obscures them, and so you need to duck into the shade to see the numbers. The engineers should have used L.C.D. (black numbers on gray background), which in any event is a terrific power saver, rather than L.E.D. (red numbers on black background). The next model, one can safely assume, will incorporate this reform (and include the missing planets). Conceivably it will do away entirely with the iron lung, and perhaps even give us the means of translating Loran sights into latitude and longitude. The Navicomp is by any standard the most comprehensively useful electronic navigational calculator made, although I have described the glorious features of the HP-41C. But as noted, the HP does not have an inbuilt chronometer—that stroke of genius is Plath’s. And, at $1,482, it certainly should be unique. A British writer estimates that its use saves one and one-half hours of paperwork per day for the navigator. Here is a suggestion for boat owners. Buy an extra iron lung ($210), keep it in your study, and take the computer home when you leave the boat. The advantages are threefold: 1) you reduce the risk of the computer’s being stolen from the boat; 2) you eliminate having to readjust the voltage from 110 to whatever you have on board the boat; and 3) for
a modest investment, you have a chronometer and a calculator for home use.

  Plath is renowned for its fine service. Although Plath-North America will make no explicit promise, the chances are reasonable that if your computer should break down, or the chronometer need adjustment, or whatever, and it would sorely inconvenience you (let us say you are setting out for Bermuda) to be without it for the two weeks required to get it to Germany, through the factory, and back; and if the computer arrives at College Park and somebody hasn’t spilled a chocolate malted milk into it—they’ll send you a spare until your own is returned. That’s the kind of people they are. I should add that although it doesn’t give you visibility in direct sunlight, the cover is a beautifully designed tinted vinyl that fits snugly over the instrument, onto the iron lung. You can keep the vinyl cover over the instrument when you remove it from the lung. The Germans are really amazing, if you don’t count world wars.

  10

  The mortal enemy of continuously accurate celestial navigation is the plain, dumb, silly mistake. Some great pedagogue of yesteryear divined this, and so he came up with—a Form. There are different forms available, depending on which navigational system you use (different systems use different sets of tables, though they are identical in conception). You are supposed to buy and use these forms (there is a nice man at Time magazine who sends out his special form to all skippers preparing to race to Bermuda). The forms are designed to protect you from making silly mistakes.

  Let me give an example. As we know, there are sixty seconds in one minute, and sixty minutes in one hour. What I have just said applies equally in measuring angles, except that one “hour” becomes one “degree” of arc, of which there are 360 in a circle, or a globe. Chicago to Chicago is 360 degrees.

  Now let us suppose that you have reached the point where you are ready to extract the difference between a line of position (LOP) on which your ship would indeed lie if the assumed position you asserted, as a necessary step in the process of discarding falsities, were accidentally accurate, and the line of position on which you in fact lie, as witness the angle you measured with your sextant.

  Here would be a typical example. Your (corrected) sextant angle (Ho)—for Height observed—was, let’s say, 35 degrees 52 minutes. Your observed (i.e., hypothetical, or “calculated”) angle (He) was 36 degrees and 12 minutes. What is the difference? Well, I would merely close my eyes (as you might do) for one quick second and mentally calculate that the difference between 52 and 60 is eight, and eight plus 12 is 20. There you have it. Right?

  Right; and it is, I swear to you, at this vulgar level that mistake after mistake after mistake is made. I am not saying that there aren’t among us a legion of practitioners who will go the whole of their lives without making an error in so preposterously simple a situation. I am saying that there are a few of us who will make such errors, and a generation’s experience with navigation establishes that I am one of them.

  The purpose of The Form (which, I confess at the outset, I do not use, but highly recommend) is to make it impossible to make such an error. Well, not quite impossible, but The Form does reintroduce every methodological seam, in order to discourage gulping down two steps in place of one. Accordingly, such a form would require you to proceed:

  Ho: 35° 52 minutes

  He: 36° 12 minutes Are the Ho and the Hc the same as to degrees? Check one:

  (yes)

  (no)

  If no, then rewrite, reducing the degrees of the higher number by one (1) and adding sixty (60) to the minutes. Your larger figure then becomes:

  He 35°72 minutes Now repeat your other sight:

  Ho 35°52 minutes What is the difference between them?

  Too much of that kind of thing and you feel like getting up and strangling the teacher. Ah, but teacher knows best. Do it that way and the likelihood that you will write down something other than twenty minutes is greatly diminished. Almost certainly you will spend less time at the navigation table. Because doing it my way you whiz through, and come up with your answer in no time flat, especially with the computers there to do most of the work, including calculating the difference between Hc and Ho—but there are a dozen other silly mistakes, examples of which would serve to make the same point (a common one is reading your watch time plus or minus one minute, or the vernier scale on your sextant off by ten minutes). Then you may end by spending an entire hour triangulating in on the error you made.

  But let me show you how thoroughly vexing the whole business is. I give a striking example—an episode the very first day out from Bermuda.

  I thought it would amuse my companions, after lunch, to play for them one of the cadenzas of the HP-97: plotting a great-circle course from one point to another, as described above in the example from Madeira to Miami—i.e., giving the course changes after every five meridians of longitude.

  I brought up my HP-97 and the printed instruction booklet, which gives sample problems, flipping it open to the indicated page, under “Great Circle Course and Rhumb Line Navigation.”

  I turned to page 02-04 and read:

  “EXAMPLE: A well-known amateur naviagator [sic] sailed from St. George’s Harbor, Bermuda (32° 23’ N, 64° 41’ W) to Horta, Fayal Island (38° 32’ N, 28° 38’ W). Plot the great-circle track at 5-degree intervals.

  Keystrokes:

  32.23 ENTER 64.41 f A

  38.32 ENTER 23.38 AB” [The outputs are then given.]

  I let out a shriek of delight. I had completely forgotten that Ken Newcomer, in making out the instruction booklet, had taken as a sample problem the route I had described in Airborne. Indeed, on June 19, 1975, we had set out from St. George’s for Faial. “Now get a load of this,” I exclaimed to my friends….

  And I keyed in the position of St. George’s as given in the keystroke instructions: 32.23, 64.41. And the position of Faial, 38.32, 23.38.

  I punched the key that would yield me the Initial Heading, and the Number of Miles.

  To my great distress, I found a significant disparity between the HP’S mileage, St. George’s to Faial, and Plath’s mileage, St. George’s to Faial. One said 2,022 miles at 067.7°; the other (Plath) 1,787 miles at 067.8°.

  There is nothing more distressing psychologically than to prepare an elaborate demonstration only to have it fail. I had had a memorable experience of this a year or so earlier—fortunately, only as a voyeur. I was to be on the Johnny Carson show, merchandising one of my books, and the guest host that night was Orson Welles. He arrived with a gleam in his eye because, sealed in an envelope from a Dutch soothsayer, he had a magic number; and he was going to demonstrate to the studio audience, and to forty million or whatever others who watch the Johnny Carson show, how that soothsayer’s secret number could be divined by a series of random mathematical operations based on data taken from the audience, as per the soothsayer’s instructions.

  There ensued the most confusing and chaotic twenty minutes since the Lord ordained that there should be light; including blackboards, Ping-Pong balls, social security numbers, drivers’ license numbers—a mix positively defiant in its complexity, but through it all Orson Welles was sustained by his vision. Finally the numbers written out on the blackboard were added together by a guest (who made an error, which was promptly corrected). It came to something like 3,482,999.

  “Now!” said Welles triumphantly to Ed McMahon, his assistant, “hand over the envelope to this member of the studio audience. Come right over here, sir, don’t be afraid! Now—” Orson Welles leaned back for his moment of triumph.

  In the Green Room, where I sat with Vincent Price and the program’s producer, I saw the producer clutching his throat.

  “Open the envelope and read the number!” said Welles.

  The man did so, with some gravity, lifting the notepaper to the light.

  “It says 212.”

  “It says what!” Welles screamed.

  “Two hundred and twelve.”

  Never was the forthcoming advertisement for H
alo shampoo more universally welcomed, by advertiser, star, and producer. When the show resumed, Vincent Price (there to promote his Oscar Wilde one-man show) was sitting opposite Welles. “Orson, how long have we been friends?”

  “Oh,” said Welles, “I’d say forty years?”

  “Then dear Orson, you won’t mind if I tell you you have laid the biggest egg in the history of television?”

  I hadn’t done quite that, but I was miserably disconcerted, and excused myself to rework the problem on the Plath—which gave me back cockily the identical readings it had done earlier, and on the basis of which we had set our course.

  So back I went to the cockpit, seeking to recapture the attention of my friends. It is wise, when you wish to move with great precision, to have someone else depress the buttons while you dictate instructions.

  “All right, Danny, St. George’s, 32.23, 64.41; Horta, 38.32, 23 38-”

  Danny looked up. “That’s not what the manual gives.”

  I looked up. Surely we had had enough problems with the coordinates of our itinerary to last one voyage.

  But there it was: a simple error. The HP people had printed the wrong longitude for Horta. They had it right (see above) under “Example,” wrong under “Keystrokes,” from which we were reading.

  I have come to know Ken Newcomer well, mostly over the telephone. And I chided him, in another book, for the mistakes in the printed literature of Hewlett-Packard, and for the opacity of the explanations Hewlett-Packard gives (although by contrast with Texas Instruments, H-P is like following a fairway). Even Plath, with its expensive machine, sends along a printed list of errors in its instruction manual, which you are given the burden of correcting, for reasons not entirely plain, since you aren’t the person who made the mistakes—you are the person who paid fourteen hundred dollars.