Land assembly & the deal

This chapter reads partial, and the reason it gives — the record carries no conveyance instrument — is exactly right. It is worth understanding why, because the corpus holds a great deal about this land and none of it is a deed.

What the record does establish

The campus is 243.0924 acres, and two independent measurements now agree to four decimal places: the metes-and-bounds legal description in the Development Agreement’s Exhibit A, and the Shelby County auditor’s CAMA polygon. [verified] That agreement retires any doubt about which ground is at issue.

It was assembled from five predecessor tracts, and the way that became known is worth recording. #1379 reconstructed the five geometrically — testing the current CAMA parcel against a 2023 statewide extract — with no instrument naming any of them. Two City instruments now recite the same five parcel numbers independently. [verified] A geometric reconstruction was confirmed by document.

And the acquisition was at least three recorded deeds, not one. The Development Agreement’s Exhibit A closes with “Being all of the lands now or formerly owned by” and names three, each to Amazon Data Services Inc., each at a different page of Official Record volume 2329 — 445, 449 and 454. [verified] The auditor’s CAMA carried only the last of the three against the consolidated parcel, so a reader working from the tax record alone would see a single conveyance.

What it does not

Everything that makes a conveyance chain a chain. Grantors, sequential instrument numbers, consideration per deed, easements — all of it lives in the Shelby County Recorder’s index, and none of it is here. The CAMA gives one conveyance date (2025-11-24) and one price ($5,621,490) [verified]; how that price distributes across three instruments is [open].

Three deeds are named for five predecessor tracts. Whether parcels -251-001 and -251-002 came in under one of the three or under further instruments not recited is not determinable from this record. [open]

The lot numbers do not reconcile, and that is a finding

Three instruments describe how the five tracts resolved into lots, and they do not agree.

[verified] for each, as a reading of that source. They cannot all describe the same object.

This study previously said the resolution was the strongest of the three because it is the enacted act. That no longer stands. The approved minutes of the very meeting that adopted it are a co-equal City record of the same act, and the City Planner presenting the legislation described it as one “that would create one new lot (#7658) from three existing lots located at the … northwest corner of Millcreek Road, Fair Road, and South Vandemark Road” — agreeing with the auditor. So the tally is 7658 (the minutes, plus the CAMA) against 7647 (the resolution’s §1).

Weigh with the resolution’s own defect in the same sentence: “to create two (1) newly-configured lot.” The numeral and the word disagree and the noun is singular. That is drafting error on the face of the instrument, which is a reason to doubt the number beside it rather than to prefer it.

Still [open] — and the thing that settles it is the recorded plat itself, in the same recorder’s index as the deeds. A smaller split rides along: the legislative record titles it “Consolidation & Right-Of-Way Dedication Plat of Survey”, the county’s legal description “COSOLIDATION & ROADWAY DEDICATION PLAT” (sic). Cite each to its source; neither is a transcription mistake.

A smaller trap sits underneath it, and it got worse on inspection. The same parcels appear in four spellings across the City’s own record:

This study previously offered [inference] that the 01/02 prefix distinguishes taxing district and the campus had been annexed. That reading is undercut: two instruments dated nine months after the annexation still say 02, so the split does not track the annexation and does not track document class either. Record it as an unresolved transcription inconsistency in the City’s own record. What is practical rather than explanatory: the auditor’s tax district is 01, so a CAMA join keyed on 02 silently returns nothing.