Most of the survey map I am working from were generated between 1825 and 1855. These are the result of a 30 year effort to produce an accurate and reasonably complete set of surveys for the territory/state of Florida. Survey's were conducted in the field by a 4 or 5 man crew consisting of a Deputy Surveyor, two chain-men, and an ax-man. These were rugged and inhospitable virgin woods. The crew was paid $4-$6 mile to do the field work. Payment computations are on the margins of the plat sheets.
Surveyors working in the field were required to leave physical indicators of township, range and section corners. They may have been an iron pipe, and a heavy wooden stake. Because of the high probability that the survey markers would be disturbed, or degraded with time, trees in the immediate area of the corner were designated witness trees. Their exact position relative to the corner was documented and recorded in the field notes. The witness trees were usually marked with a cat-face, or chevrons, to positively identify them. Some of those marked trees survived beyond 100 years of age, and a rare few made it to 200 years. One tree, was documented as germinating several years before the constitution was signed. It was inadvertently cut down around 1981.
The initial result of the field surveys was a hand written note book, giving dates, measurements, compass headings, etc. This information was being calculated in the field using basic equipment, which was state of the art for the day. The completed field notebooks were returned to the land office responsible for that part of the state. There, the numbers would be verified, and a drafts-person would generate a visual survey document. That would have to be rechecked for errors, before being approved by the surveyor general, or one of his deputies.
Comparison of the finished sheets, to current USGS topographical maps, suggests that the drafts-persons may have used a wee bit of literary license with some of the drawings. The features may well have existed, but their placement might not always be spot on. So beware,