A while ago, there was some discussion of the accuracy of 19th century chaining. Here's an example from a series of surveys run in the vicinity of Wimberley, Texas in 1847, back when surveyoring meant having to contend with the armed neighborhood watch committees of the Texas Indians in possession of the land.
These are all the distances between corners of nominally 640-acre tracts surveyed off in 1847 as those corners were located from original evidence of that 1847 survey, mostly in the form of bearing trees. The distances in the "1847 Record" column are the lengths that the 1847 surveyor reported in the metes and bounds description of the tracts he prepared and filed in the General Land Office.
[pre]
From-To Actual Grnd 1847 Record
1288-188 1581.02 vrs 1582 vrs.
189-188 2294.73 2283
188-187 2298.18 2283
186-187 1590.49 1582
187-209 1589.27 1582
[/pre]
The terrain that the lines ran over was generally without steep slopes and likely without dense brush.
So, what do the above differences say about errors in the 1847 chaining? First off, there appears to be a more or less systematic error in play in the chained distances reported by the 1847 surveyor:
[pre]
From-To Actual Grnd 1847 Record
1288-188 1581.02 vrs 1582 vrs.
189-188 2294.73 2283
188-187 2298.18 2283
186-187 1590.49 1582
187-209 1589.27 1582
--------- ------
9353.69 9312
[/pre]
So, what that would mean is that the average scale factor that relates the 1847 record to actual is
9353.69/9312 = 1.00448
The statutory Texas vara is 33-1/3 inches, so that means that the 1847 surveyor was working with what in effect was a vara that was:
33-1/3 varas x 1.00448 = 33.4827 inches, which is nominally 33-1/2 inches.
Assuming that there was a systematic difference in chaining in the 1847 work, the obvious next comparison would be to scale the 1847 record distances by the chaining factor of 1.00448 and see how well invidual lines fit.
[pre]
From-To Actual Grnd Rec x 1.00448 Diff.
1288-188 1581.02 vrs 1589.09 vrs. -9.07
189-188 2294.73 2293.23 +1.50
188-187 2298.18 2293.23 +4.95
186-187 1590.49 1589.09 vrs. +1.40
187-209 1589.27 1589.09 vrs. +0.18
[/pre]
That is very impressive.
So what did they measure with?
I have no knowledge of Texas surveying except from what I have read here,
Was there a vara "chain"? Like 100 varas in length or something similar...
Or did they use a chain and convert to varas. Most of these Texan surveyors seemed to have driftedf into Texas at some point.
> Was there a vara "chain"? Like 100 varas in length or something similar...
The standard chain was 10 varas in length, nominally 27.78 ft. from handle to handle and so was comparable to a 2-pole chain.
The problem with many surveys made in the 1840's in Texas may well have been that there were still chains in use that were regulated to the varas in use during the Mexican colonial period. It was one thing for the Commissioner of the Texas General Land Office to send around a letter instructing all of the county surveyors what length a vara was to be and quite another to make them toss out all of the older chains they already had and that were useful for following the lines of older grants run during the colonial period.
The surveyor who signed the field notes of the 1847 surveys was a fellow named Bartlett Sims who had been active in the area during the colonial period. It wasn't his first rodeo.
Ray Skelton was a master with words. One passage in his book really stuck with me. Of course I'm paraphrasing, but he noted that when you were evaluating or retracing another surveyor and his measurements, you should "adjust your chain" to find or set the tract back in.
While I'm sure the man didn't mean to cut ones chain to fit, his metaphor for evaluation and retracement has lead me to make some form of scale factor on every job that I've done since I read that circa 1999, in evaluation of their "chain" vs. my "chain".
:good:
Thanks, Kent. We hope to be out in the woods very soon. So many stones to find, so little time.
Kent,
I had some free time today to really play around with your post. While I agree there probably a systematic error there seems to be something else in play in your line from points 1288-188.
[pre]
From-To Actual Grnd 1847 Record Scale per line
1288-188 1581.02 vrs 1582 vrs. 0.99938
189-188 2294.73 2283 1.00514
188-187 2298.18 2283 1.00649
186-187 1590.49 1582 1.00537
187-209 1589.27 1582 1.00487
--------- ------
9353.69 9312
[/pre]
I come up with 2 possible reasons for the difference in line 1288-188 and the rest of the lines:
- Line 1288-188 is a closing line so other factors such as angles may be coming into play.
- Or the original surveyor tallied one to many chains, assuming they were using a 10 vara chain. If you use a record distance of 1572 vrs for line 1288-188 the scale factor for the line calculates as 1.00578 right in the mix of the scale factors of the other lines.
Just 2 possible guesses based on the data, any chance either may have happened?
