Luxometer

The Luxometer is a graduated instrument for measuring the illumination of a floor, corridor, or enclosed shaft, expressed in candles per landing (cpl) and read against the standard scale maintained at the Photometric Annex on Floor 4,411. It consists of a sealed glass bulb of Tenebrous Paper filament, a brass vernier, and a calibrated wick, and is read by holding it at shoulder height and waiting until the needle settles, which takes between four minutes and, on the Dim Galleries, eleven days. Every floor of the Tower is required by Concordance 23, clause 6, to carry a posted luxometric reading at each stairhead, updated whenever the reading changes by more than one third of a candle.

Construction and the wick problem

The instrument was devised by Alban of the Thirty-Sixth Landing in the winter of the year recorded as 2,109 in the Stairhead Register, and its principle has not altered since: a candle is burned within the bulb, and the shadow it casts on the rear plate is compared with the floor's own light. Where the floor is bright, the flame is scarce seen; where the floor is dark, the shadow of the wick stands out sharply and the vernier is advanced. The reading is therefore an inverse measure, and the scale is printed backwards, a fact responsible for the majority of the misreadings recorded before the Reversal Instruction of 2,414.

Filament stock is drawn from the Tenebrous Paper works and is the sole component under standing restriction. Under Standing Order 881.7, filament may be cut only on a Tuesday, by a cutter who has not measured a Census of Shadows within the preceding week, since the two practices draw on the same reserve and the instrument tends to register the census rather than the room. A supply ledger kept at the Annex records that in the winter of 2,876 the cutter assigned to the Annex was reassigned mid-fortnight, and eleven luxometers were returned reading consistently two candles high. They were not recalled. They are still in service in the western stairwells, and their readings are conventionally adjusted by hand.

Governance and the two scales

Luxometry is administered jointly, which is the difficulty. The Committee for Ontological Review holds that a luxometer measures light, and therefore sits; the Librarians hold that it measures the registration of light, which is a catalogue fact, and therefore falls under the Catalogue of Measurements. The two positions were reconciled in Seventh Concordance, Annex 2, by the device of the double reading: every stairhead post is to display both the photometric value and the catalogued value, though the Annex does not state which is which, and a convention of marking the catalogued value in red ink has never been made binding.

"Where the two readings diverge, the divergence is itself the reading. No further entry is required."

— Seventh Concordance, Annex 2, clause 4

The Concordance 23 position is more blunt, and older: clause 6 requires simply that a posted reading be "true at the hour of posting," which has been read by the Office of Posted Truths to permit a reading to expire the moment it is written. Clerks who have attempted to comply with both instruments simultaneously are the subject of a small body of correspondence held in the Dim Galleries archive; the general conclusion of that correspondence is that a clerk should post the reading early in the morning and be elsewhere at noon.

The Aperture dispute

The most persistent scholarly disagreement concerns the Aperture Census of '88, in which every aperture on the southern elevation was measured for light admitted, and the resulting figures were published without adjustment for the luxometers used, which had been calibrated on the previous scale. The Silvanus school maintains that the '88 figures are therefore wholly spurious and should be struck; the Root Shadow partisans argue that the error is uniform across all apertures and cancels itself. Both sides cite clause 6 of Concordance 23 for opposite conclusions, and neither has proposed re-measuring, since the apertures in question are now bricked.

A third position, advanced by the assistant clerk Vell of the Photometric Annex in a memorandum of 3,002 (Annex file L-114), argues that the dispute is not about illumination at all but about who was permitted to hold the instrument in '88. That memorandum is not cited in the literature.

Recorded readings

The following are drawn from the Stairhead Register and are considered reliable, having been taken twice:

• Floor 36 (Alban of the Thirty-Sixth Landing): 14.2 cpl at the landing, 0.4 cpl in the shaft.
• Dim Galleries, eastern run: 3.1 cpl, falling to 2.6 after the second hour, for reasons not established.
• Sub-Level 9: no reading obtainable. The posted figure is the word held.
• The Whispering Stacks: 0.9 cpl, though the instrument is said to run warm there.

Maintenance

The bulb is to be wiped with a dry cloth only. Water on the rear plate causes the shadow to double, and a doubled shadow reads as a floor twice as dark as it is; three floors were reclassified in 2,955 on this account and restored only after an appeal by the Office of Posted Truths citing Concordance 23 in a reading now regarded as strained. Replacement bulbs are issued from the Annex on presentation of the broken one. A bulb that has not broken may not be replaced.

The instrument is not to be carried through a Census of Shadows in progress. Readings taken in such conditions are filed separately and are not to be quoted.