California Surveyors Forums

I'm not sure I quite follow the precise wording given in (at least) three different GLO manuals in the 19th century, namely the 1855, 1864 and 1881 Manuals*, all of which are applicable to California surveys. The wording in all three is very similar regarding the topic, and I'll just quote from the 1855 Manual, which states:

"EXTERIORS OR TOWNSHIP LINES.

The principal meridian, the base line, and the standard parallels having been first run, measured, and marked, and the corner boundaries thereon established, according to instructions, the process of running, measuring, and marking the exterior lines of townships will be as follows:

Townships situated NORTH of the base line, and WEST of the principal meridian.
Commence at...the southwest corner of T1N R1W., as established on the base line; thence north, on a true meridian line, four hundred and eighty chains, establishing the section and quarter section corners thereon...thence...establish the corner of Tps. 1 and 2 N Rs. 1 and 2 W.; thence east, on a random or trial line, setting temporary section and quarter section stakes, to [near the NE twp corner], where measure and note the distance at which the line intersects the eastern boundary, north or south of the true or established corner. Run and measure westward, on the true line...establishing the section and quarter section PERMANENT CORNERS on said line. Should it happen, however, that such random line falls short, or overruns in length, or intersects the eastern boundary of the township at more than three chains and fifty links distance from the true corner thereon, as compared with the corresponding boundary on the south, (either of which would indicate an important error in the surveying) the lines must be retraced, even if found necessary to remeasure the meridional boundaries of the township, (especially the western boundary,) so as to discover and correct the error; in doing which, the true corners must be established and marked, and the false ones destroyed and obliterated, to prevent confusion in [the] future..."

The part of this that throws me is: "Should it happen, however, that such random line falls short, or overruns in length, or intersects the eastern boundary of the township at more than three chains and fifty links distance from the true corner thereon, as compared with the corresponding boundary on the south..."

It's clear from this that the error limit of the NE corner of the twp, in the N-S direction, is +/- 3.5 chains, relative to the position of the NW twp corner, but I can't see that this wording gives any error limit in the E-W direction, i.e. in the allowable length of the twp N boundary line. Surely, in almost every case, the random line is going to "fall short, or overrun in length" that of the S boundary of the township, to some degree, given the tools of the time and the realities of terrain and vegetation. In other words, there's an error limit provided for the twp NE corner location in the N-S direction but none for the E-W direction.

Or, is the error limit in the E-W direction in fact 3.5 chains also but they just worded things poorly? The apparent problem with that explanation is that if so, then they worded it poorly in three consecutive instruction manuals and nobody bothered to change it.

If taken verbatim, the twp line surveyor apparently would have to re-survey and re-mark up to 12 miles of pre-existing twp line, lines that he very well may not have run, and which could very well be the direct cause of his random line not being exactly 480 chains long, and/or falling > 3.5 chains N or S of the NE twp corner. This seems like an utterly preposterous penalty to have to pay for somebody else's mistakes; if that were the case I could well see why he might just fudge his numbers to make the N boundary line appear to meet specifications.

So this is puzzling to me. What am I missing here?

* Using the versions included in White (1983) A History of the Rectangular Survey System.

This section does seem to need the benefit of a comma. While yes, the 3.5 ch limit would seem to apply, taking into account the anticipated east-west misclosure due to convergence of the meridians at the latitude being surveyed. From a practical standpoint however, it’s probably moot since likely you’re not going to be establishing an original township boundary. The rule of the land is original monuments control, regardless of the amount of error actually contained in the survey. Our job is almost exclusively one of retracement and trying to best place error where it can be proven to have occurred. I say almost exclusively since I have been involved in surveys for the BLM where the powers that be (all owners plus cadastral chiefs), where the original plat was cancelled and the township was re-platted as tracts based on limited recovery of original monuments and protection of bone fide rights. However, we operated under new special instructions with a much tighter tolerance than the old 3.5 chain rule.
Maybe there were some original lines re-run due to misclosure way back when. My experience retracing them has shown they left their errors standing. I see a larger amount of error in many sections, let alone an entire township. Original monument position controls over distance or direction.

Thanks Steffan, I much appreciate the reply.

Yes, I imagine a new township boundary line survey would be about unheard of now except in some unique situation. I'm just trying to wrap my head around how much error in a line, in the X and Y directions, would have been tolerated in those days before a given Surveyor-General would be likely (or even required) to force a surveyor back into the field to fix problems. The more stringent the standards (and strict an S-G in enforcing them), especially if unreasonable relative to the tools and land conditions of the time, the greater may have been the incentive to fudge the numbers in the official field notes.

In reading various Manuals of Instruction, I'm not finding much guidance on the issue. The only limits I can find reference to therein are the 3 to 3.5 chains per township line and the 0.8 to 1.0 chains per section line, which are presented as an absolute upper limit. On all twp plats done by surveyors associated with Benson's ring that I've looked at, their recorded distances and angles fall under these limits (of course).

But the error given across a series of lines--such as sets of 5 to 6 consecutive section lines when moving northward from the S boundaries of townships--do not appear to accumulate, which they seemingly should in theory, given that the positions of section corners on such lines are never corrected after running random and true, but instead are given as lying exactly 80 chains N of the one S of it, and always on a true north bearing. This cannot of course be the case. There's some unknown and undeclared per-mile error rate in the X and Y directions, and since the corner positions along a line are not changed once set, that error has to accumulate, as a random walk--the error in the random lines running eastward, closing each section, has to be expected to increase in moving from S to N through a twp. This essentially never happens in Benson Syndicate twp maps and I'm starting to wonder about how common it was in general, or conversely, whether I've missed something.