Medial entorhinal cortex
The block, to scale
Two reconstructed cells sitting inside the imaged volume, at true relative size and in the positions they actually occupy. Drag to orbit, scroll to zoom.
Everything here is placed by measurement, not by eye. The wireframe is the segmented block at 2031 × 1338 × 447 µm, taken from the segmentation's own bounds at 16 × 16 × 45 nm per voxel. The two cells are real reconstructions pulled from the chunked graph, and each one sits where it actually sits: their bounding box centres are 148 and 417 µm from the block centre along the long axis.
Turn it edge on and the thin axis does the explaining. The block is only 447 µm deep, and both cells span about 317 µm of that, so they very nearly fill it. The green cell reaches 1312 µm across the block, roughly two thirds of its width, almost all of that a single long process. Arbor that leaves the block is simply not in the data, so what you see is cut off by the edges of the volume, not by the cell.
These are two cells out of tens of thousands. They show scale and position. They say nothing about how the rest of the block is populated.