I'm in the somewhat early stages of retracing some old original surveys of Ohio's Connecticut western Reserve.
One thing I noticed quickly was that chained measurements were, on average about 50 or so feet long per mile.
That is, that a report of 80 chains(1 mile), results in a 2012 measurement of somewhat less than 81 chains.
There's no notation about extra links and I don't personally subscribed to some universal addition of an extra link in order to not being able to report a relatively accurate measure. In other words, in general, I believe that surveyors(even in late 1700's), made an attempt to "accurately", measure distances.
I'm lucky inthat other surveyors(government) were charged with "re-measuring", along the southerly 120 mile long line and to note the relative locations of found monumentation . . . another reason to believe that there was no "accepted", general allowance to bastardize measurements by adding an extra link, for "good measure".
In any case, each of the subsequent surveyors, although showing differences in the previously reported measures, generally indicated their own "lengthened", dimensions of the average chain. The later surveyors, for the most part appear to have measured more "Accurately", but still exhibited a lengthened chain of about 20-30+ links or so per mile on average, while calling the original survey off by 20 or so links lengthened measure by the original surveyor in 1797.
Another poster from the Logan Ohio area has also posted here that measurements in the 7 ranges seem to suffer similar lengthened measures.
I'm kinda shocked and don't understand how these lengthened measures have occurred.
I'm fully aware of the loops in a chain stretching, but I perceive that any surveyor who takes some degree of care in measurement would have a way of knowing . . . at least from time to time . . . how his chain is measuring. Especially the government surveyors who were specifically charged with keeping a chain specifically for a stated regimen of checking the working chains.
With that being said, measurements of one of the government surveyors who followed the 1797, actually exhibited the continuous lengthening that would occur with a stretching chain for while, then all of a sudden begin to exhibit a more in line "accurate" measure, until the measurements would again exhibit a lengthening.
Problem is that even the more accurate measurements(before the noticeably longer measures), exhibited a considerably lengthened measurement.
In other words . . . it really does appears that measurements of a chain's length(by 3 different surveyors at 3 different times from 1797 to about 1808), was generally resulting in more than a chain's length when compared in 2012 terms.
Any thoughts on this . . . other than the "added link" concept?
check with 'carl zeiss', i believe he was the one that posted earlier of the same
If the measurements are long, then the chain is "short".
Didn't the original surveyors "calibrate" their chain? I thought that was standard procedure. Maybe search for those notes.
As I am sure you are aware, in that time period it was common practice for a fairly large percentage of the field surveyors to add a "for good measure" amount to every length they measured. It seems you have found one of these areas.
Have you analyzed the discrepancies by separating the different surveyors to determine a consistent coefficient used by each surveyor?
B-)
>If the measurements are long, then the chain is "short".
No. If the original distance was longer than they intended, it is because their chain was long.
If you get a larger number than true distance it is because YOUR chain is short.
Their stretch seems to be more than I would expect from reading about the mid 1800's practice, but I don't know much about your era. Were the chains of similar construction back then?
If you convert 50 ft/mile into thousandths of an inch per ring in the chain, it doesn't seem quite so huge, but still strikes me as excessive. The sudden reversion to more accurate measure does seem to indicate they standardized once in a while.
How consistent is this chaining difference? If isolated could be dropped links?
If the measurements are long, then the chain is "short".
Not really . . . if a chain is long, then an unconstrained 80 chain long measure will be more than 80 chains long. That is . . . if the chain is 67 feet long, then a measure of 80 chains(to actually set points A & B), will actually result in a measure of 80 chains + the extra 80 feet.
If the measure is constrained(such as a re-measure of an 80 chain length from point A to point B), then a measurement from point A to point B, with a long chain will result in a reported measure(for the distance A-B), that is shorter.
Since the original measurements were long(to the tune of about 40-50 feet per mile), I would expect subsequent measurements(by later surveyors), to report the previously set(re-found/re-discovered), monuments, "reported" to be further downline than originally reported.
This is the case in the 1797 survey and the 1799 and 1808 re-surveys of the line.
The 1799 & 1808 surveys both appear to report a mile(80 chains) as measured in 1797 to be more than a mile.
Problem is, both the 1799 and 1808 re-surveys also appear to suffer a similar lengthening as did the 1797 survey, but to a lessor degree.
This makes me believe that a well made, well maintained chain might have generally resulted, by some methodology, in a longer than actual measure.
I'm trying to understand what this quirk(or something), in methodology might actually be the cause for the difference.
I'm willing to bet that this "good measure" concept was used in private surveys.
I seriously doubt that this was a common practice for original surveys and less so for government surveys, let alone for "re"-surveys of an original survey by government surveyors.
If the measure is constrained(such as a re-measure of an 80 chain length from point A to point B), then a measurement from point A to point B, with a long chain will result in a reported measure(for the distance A-B), that is shorter.
And a short chain will result in a reported measure that is longer.
Keep in mind, we're talking about three surveys. The original survey of 120 miles, a resurvey by government surveyors a couple years later and another re-survey by government surveyors about 9 years later(than the original).
The government surveyors were governed by standards set up by the government as to what type chain and methodology to ensure more accurate measures.
The original survey was not performed by a government surveyor(?). It was a survey to demark Connecticut's claim to that part of Ohio that the state of Connecticut claimed from King George and that portion of land the new federal government allowed Connecticut to keep.
This land was being surveyed for the owners(bought from the state of Connecticut).
So, as I see it, this was, more or less, a "private" survey.
But the person who actually measured the 120 miles was supposed to be highly qualified and became . . . if I remember right . . . the Surveyor General for the U.S., for a while.
In other words, this guy(Seth Pease), was not some duffus with a compass and chain. He was a person who could be expected to know what he was doing.
The instance of measurements coming out so much longer(than "recorded"), made me wonder about him.
The fact that the subsequent surveyors were actually closer to his measurements than to the measurements(in 2012), and that ALL THREE surveys generally resulted in longer measures, lead me to believe that, in those days, a measure of 1 chain could, with some quantifiable standard of care, result in a more than 66' measure.
I want to understand what it was that could cause measurements made with what appears to have been proper methodology(for the time), to result in a longer measure.
If the measurements were slightly shorter(or very nearly the same), I wouldn't have given the difference much thought.
It's the lengthened measurements that has me scratching my head in wonderment.
Granted . . . and I will review the subsequent "re"-surveys with this in mind to see how a short chain(in 1799 & 1808), might agree or disagree more(or less), than the original.
But, we're talking about the "original survey", and the original survey was not constrained in any way.
The subsequent re-surveys generally report distances between the original distances and the distances from 2012.
This would mean that even if the later re-surveys call for differences(or resultant errors), from the 1797 survey, that they too are resulting in a chain being longer than 66 feet.
Both, the 1799 & 1808 surveys are generally n agreement that the 1 mile measure of the 1797 survey is more than a mile. In 2012, I find that the 1797 surveyed measurements are longer than the 1799 & the 1808 re-surveys, but that all three measured more than a mile for each mile measured.
Had the 1799 and 1808 surveys been measured with a short-chain, they would've reported the 1797 survey being longer yet . . . not shorter.
Had they used the exact same methodology and all things were equal . . . other than the chain length, the 1799 & 1808 surveyors would've had shorter chains(very possible considering sag and all).
BUT, the 1799 & 1808 measurements are also longer than the measurements derived in 2012, which means, even if they used a shorter chain, the chain was still longer than 66 feet.
All this might be possible . . . all I'm hoping for is the best definitive answer that will(or might), explain the how and why.
rough example just FYI
These are not actual measures. These are just representative
distances that show how the numbers are comparing to one another
ORIGINALLY
1797 reported: 80.0 chains
RE-SURVEYS
1799 reported: 80.4 chains
1808 reported: 80.5 chains
2012 DATA FROM
GPS & SURVEYS
2012 derived : 80.8 chains
No added link but thinking back to the educational level at the time, if may be something as not remembering or something like that. All kinds of variables come to the surface when you consider that many people couldn't read and write until the 1930's even.
rough example just FYI
That is typical of what I will often see. SOP in these parts, .4chs long per 1/2 mile is about right on.
John, Have you seen
any possible 'Temp correction' or the lack thereof, in reference to the time of the year?
I have read GLO notes where the surveyor indicates that he adjusted the length of his chain usually by making it longer to fit a previous range or township line. 40 links per half mile seems to be about it. That's an extra link per chain length. This is in the 1870 to 1900 time frame.
I'm working on one right now where I've found original corners from 1895 for about 6 miles north/south, all reported to be 40 chains per half mile. Average chain length from my measurements is about 66.7 feet. A high percentage of the GLO survey's in my local area are about link long in the chaining. Some of the notes tell you that but some don't. Then I have one township in the middle where the chain was held and the lots on the north side all have 4 to 5 extra chain lengths and closing corners along the range lines where they ordinarily wouldn't be. I suppose this deputy just couldn't stomach adjusting the length of his chain from standard.
Since the original surveys control it's just what it is. Once I figure out about what the original chain length for that crew during that survey was it makes it easier to locate things.
Most of the work I do in Western Illinois was surveyed around 1817. I would also agree that in my area that a half mile is about 0.4 chains long. I've surveyed in this area for close to 30 years and its tough to say the number is consistent, but it is almost always long. A distance of 2650 to 2680 seems to be a common range for a half mile in my area.
My personal opinion is, and I've seen other posters mention it, that they just added a little extra to make sure it was at least a half mile. I have no proof of this. It seems like going up and down and hills, not running a straight line, would make all the distances short. The only other explanation is that the chains wore and were quite a bit longer than 33 feet.
The idea of a long chain makes sense to me except that I always figured government surveyors would have made all attempts to measure for a "true" accuracy and not to pay heed to giving extra allowances.
If everyone(even the government), did this(use a long chain), I would think that the possibility wouldn't be just rumor, but rather a definitive, verifiable fact.
ohhhh . . . my head hurts
> The government surveyors were governed by standards set up by the government as to what type chain and methodology to ensure more accurate measures.
While this may be true, it is not guaranteed and set in stone that every gov't surveyor followed the standards to the letter each and every time they laid the chain on the ground. They generally took short cuts (measuring on slope and estimating the horizontal correction for example). Knowing they were taking short cuts they probably compensated to "error on the long side" of things, sorry but that is just the way it was. Ever wonder where the saying "close enough for government work" came from? 😉 They were probably paid by the mile, if they had been paid by the hour I would think you would see more careful and precise measurements (heck they'd probably still be running the lines!!)
Just be happy the "error" was reasonably consistent and you can find the original lines and corners.
There is a pretty good chance they just didn't care. The way we survey today, to the nearest 0.01' would probably blow the old surveyors away. At the time, the ground was mostly open, seemed endless, and was going cheap.