Date: Mon, 6 Nov 2000

Peary should have taken some kind of a sight 50 miles short of the pole. Actually, 89-25 would have been a decent place to do this, and depending on what he did while he was staying up all night, this could explain his accuracy. Maybe I need to get a mercury horizon and some electric socks and get someone to fly me up to 89-25 and see what can be done by eye.

I have Molett's book, and this kind of approach absolutely can be used at least up until the point the sun does not set (March 26 in Peary's case). Also, Bartlett wrote to Grosvenor during the Helgeson brouhaha that "we checked the compass courses between headlands and set our course north. We went out and watched the sun rise and set..." or something like that. I don't think Bartlett threw in the stuff about the sun rising and setting as a casual comment about pretty colors or something. I think he was referring to the fact that the time of sunrise and sunset could be noted and the midpoint between these times is the time the shadows will point due north (actually, slightly to the right of due north, when the sun's declination is increasing, but close enough).

This is why I am pretty well convinced that on March 28 or 29 (not sure which), when Bartlett took his sounding at latitude 87-15 (no bottom) he was over the canyon that lies just to the west of the 70th meridian, rather than (the other possibilities, based on ocean depths) way off to the west or off to the east. My only problem with Molett is that he is adamant that using the time of the sun's culmination does not work. It is pretty clear Peary thought it did work, since he hired Hastings to check the accuracy of the method. Also, I have done it many times, and it does work.

Molett did all his navigating from an airplane, using a bubble sextant. I am quite sure you could not do the culmination bit with a bubble sextant under the best of circumstances, and in any case not on an airplane (accelerations of the plane get the bubble out of its straight up position, for one thing, and for another, just the fact that the airplane is moving at a high rate of speed would introduce errors in the time of culmination).

The navigation article from 1907 is referred to in Roland Huntsford's book about the race between Scott and Amundsen. His book also refers to an article that claims that Amundsen took no sights for longitude prior to reaching the very near vicinity of the south pole. We cited that article in our report, and Ted Heckathorn came out of the woodwork claiming that the norwegian version of Amundsen's book includes a longitude sight. I never followed up on this, but it would be interesting to see if Heckathorn ever published anything on this. (Boyce Rensberger, Rawlins' pet reporter, gave it full coverage in the Washington Post, so you would think it had to have been published, but who knows?) Maybe Bert Stafford knows something about this. I think Heckathorn is an acquaintance of his.

Rawlins probably knows about this, but he never looks at Peary's navigation from the point of view of navigational methods known at the time. He assumes since nothing has changed much with the stars and planets over the last hundred years, the same is true of navigation. Rawlins has made a number of errors indicating his ignorance of practical navigation, and a number more indicating that he didn't appreciate the limitations on the methods Peary used.

(1) In the famous "slip" slip up, Rawlins expressly states that "Peary strained to interpolate his sextant vernier beyond its normal accuracy." (or something like that. I often use quotation marks when I don't have the exact quote in front of me, so don't hold me to these.) This Rawlins did so that he can claim that numbers that included minutes and seconds and half seconds could be interpreted as angles, rather than times (which is what they were). Peary's sextant could be read to the nearest 10 seconds, by using a magnifying glass to look at the vernier scale to see which pair of finely engraved lines matched up. (As far as I know, 10 seconds is the most precision ever built into vernier on limb type sextants.

Some of the "endless tangent screw" type, with a vernier on a drum on the tangent screw go down to .1 minutes or 6 seconds.) The idea that anyone could interpolate a vernier scale is utter nonsense. It is absolutely impossible. Much less can it be interpolated to 20 times its normal accuracy. This is like saying you wanted to note the time very carefully, so you interpolated the time between seconds on a digital watch and estimated the time to the nearest 20th of a second. It simply cannot be done. What's more, there would be no purpose in trying to read a sextant to this degree of accuracy.

Peary himself said that with uncertainties in refraction, he would expect possible errors in the range of 5 miles (5 minutes of arc). In the best of circumstances, a navigator expects to be within a mile (1 minute of arc). In fact, some of the very best sextants ever made only have a scale to the nearest minute, allowing the user to interpolate (let's say to a quarter minute) if he is really anal. The advantage of this type of scale is that it is easy to read and avoids the risk of errors of several minutes that can occur when one gets a bit confused reading a vernier scale. Peary would not be so crazy as to try to work out his position to 1/2 second (about 50 feet).

(2) Rawlins assumes that single noon observations, such as Peary's first one at Camp Jesup, Bartlett's sight, and Peary's 1906 farthest north are "imperfect" because they might not have been taken at culmination, and refers to multiple observations as "culmination sets." This all sounds fine, but in either case, the sun is observed continuously until it has started to set. (Bartlett described his method, and a Cornell professor described the method Marvin used.) The multiple observations are not taken to be sure that culmination has occurred (that is ensured by the continuous, although unrecorded observation in either case) but to eliminate errors that can creep in due to the prismatic effect of the glass cover for the artificial horizon, if the two surfaces of each plate of glass are not perfectly parallel and possible inaccuracies in the estimate of the sun's apparent semidiameter. These are good improvements, but would not be expected to make a difference of a minute or so.

(3) Rawlins keeps talking about "Sumner lines," which are lines of possible positions for a given sight. These are named after Charles Sumner, who, in about 1832 or so, worked out two different longitudes for a single sight, based on two different assumed latitudes, and then drew a line connecting the two lat-lon positions.

Marc St. Hilaire published his "new" navigation method in the 1870s, allowing a line of position to be generated more efficiently by calculating the azimuth and drawing the line of position perpendicular to the azimuth through the single lat-lon point. This was followed by tabular methods that allowed the navigator to assume an even degree of latitude and use pre-calculated longitude (or, more accurately, local hour angle of the sun) and azimuth to plot the line of position.

All of these developments were extremely valuable to sailors, who were always trying to get what information they could about their location as they moved along. They were relatively unimportant to surveyors, who were content to first determine, with great accuracy, their latitude and then do a longitude sight, with the calculation based on that latitude. This is the type of calculation that Peary was trained in and Mitchell and Duvall, as well.

Rawlins ridicules their complex calculations that could have been approximated by graphing easily calculated lines of position. There is nothing to suggest Peary was trained in, or thought about, lines of position. As noted, position line navigation becomes very easy near the pole, because one does not need pre-computed azimuth or altitudes. The altitude of the sun at the pole is simply its declination and its azimuth is equal to its local hour angle, both taken from the almanac for the GMT of the sight. What is more, this method is quite accurate surprisingly far from the pole. Certainly at Bartlett's camp, this would have provided a very easy method of getting a precise location. Unfortunately, Peary was not aware of this. He was not a theoretician when it came to navigation, and theoreticians did not devote much thought to navigating at the pole prior to the invention of aircraft.

(4) Rawlins ridicules Peary's statement that he didn't use his theodolite, but would have done so if the sun had been higher. He says anyone who has ever seen a theodolite realizes how foolish it is to say it cannot be used at low altitudes of the sun. (Has Rawlins ever seen a sextant?) Why would Peary say something foolish, for no apparent reason. Answer: He didn't say the theodolite could not be used for a low sun, he just said he would have used it if the sun were higher. On ice, Peary preferred the sextant and mercury horizon, because the theodolite tripod legs would melt into the ice at different rates, constantly throwing the instrument out of level. The mercury horizon was self leveling. However, anyone who has tried taking sights with a sextant and artificial horizon lying prone knows that it gets awkward if the sun is higher. At some point, the observer must change to a sitting position, but there is a point in between which is awkward. Also, the sitting position does not provide the stability and relieve the arm strain from holding a heavy sextant (typically about 7 pounds) for a long period to observe the slow rising and culmination of the sun.

In any event, it is foolish to make a big deal out of this remark of Peary's, but Rawlins plays it up as if he has caught Peary in a big lie.