Ephemeral Minerals

Ephemeral Minerals are those mineralogical substances whose primary attribute is a demonstrable period of material non-existence. They are catalogued, graded, weighed, and audited in precisely the same manner as ordinary minerals, save that for part of each cycle there is nothing to weigh. The definitive registry of these substances is the Table of Ephemeral Minerals, the annex to the Standard Taxonomy of Non-Existent Minerals, and the foundational typology for the major classifications of ephemeral phenomena at large. A mineral is not ephemeral because it is rare, fragile, or quickly consumed; it is ephemeral because it is, for a stated interval, absent, and returns afterward in the same condition in which it departed.

Definition and criteria

The distinguishing criterion is recurrence of absence. A mineral that crumbles, dissolves, or is worn away has merely ceased to exist once, and is filed under ordinary attrition. An ephemeral mineral exists, does not exist, and exists again, and does so on a schedule that can be written down in advance. Somnic Chlorite, the best-studied example, is absent for intervals of between forty minutes and eleven days depending on the sample, and the interval is remarkably stable for any given specimen — a property that has made it indispensable to the calibration of the Ephemeral Gradients system, which confirms the absence of ephemeral minerals during their absent phase.

Absence is not the same as invisibility, and the distinction has cost the Tower considerable expense. An invisible mineral can be detected by its displacement of air, its weight on a pan, or the shadow it casts in strong light. An absent mineral does none of these things. The standard test, set out in ephemeral audit procedure, is to leave the specimen pan empty and undisturbed and to record what does not appear on it.

Classification within the Table

The Table divides ephemeral minerals into four orders according to the duration and regularity of the absent phase, from the brief intermittent class (absences under one hour, repeated at least daily) to the long-cycle class (absences exceeding a year, of which only two examples are known and one is disputed). The Table further records, for each entry, the substrate it leaves behind during absence — usually a faint residue of dust, occasionally nothing at all, and in the case of the laminated Ephemeral Mica, a residue that can be peeled.

The Table is held jointly by the Committee for Ephemeral Cohesion and the Office of Ephemeral Cartography, an arrangement that has never been described by either body as satisfactory. The Committee formally adopted the Table upon its constitution in Year 12,477 of the Tower Record; the Office maintains that custodianship was assigned to it by an earlier instrument and that the Committee's adoption was therefore a courtesy rather than a title.

The custodianship dispute

The dispute turns on the wording of Point 12, which assigns to the Office "the care of all substances whose situation cannot be relied upon." The Committee reads "situation" as spatial, and therefore as referring to the stacks and their logistics; the Office reads it as temporal, and therefore as covering any mineral that is sometimes not there. Both readings are defensible, and both have been upheld.

"Where two readings of a Point are each defensible, the earlier reading shall be preferred, unless the later reading has been acted upon for a period exceeding the memory of the office acting upon it."

— Fourth Concordance, Annex 9

As the Office has acted on the temporal reading since before any living clerk can recall, and the Committee's reading is demonstrably the earlier, the Annex establishes nothing whatever, and the dispute proceeds. The Bureau of Ephemeral Holdings has twice attempted mediation. Its first report recommended that the Table be copied and each office hold one copy; both offices rejected this on the grounds that a copy of the Table is not the Table, and that a mineral listed in a copy may not be absent when consulted.

Practical difficulties

Ephemeral minerals are awkward to store. The vaults of the Cartography of Ephemeral Stacks reserve a separate shelf for each entry, and the shelf is kept clear whether or not the specimen is present, so that an absent mineral is never disturbed by a clerk reaching for something else. Shelves are marked with the date and hour of the next expected absence. A specimen removed from its shelf during an absent phase cannot be returned to it, because there is nothing to return; the practice is to wait.

The Ordinance on Ephemeral Inks arose from precisely this difficulty. In the winter of Year 12,509 a clerk of the Office of Ephemeral Affairs recorded the return of a Somnic Chlorite sample in an ink since found to share the mineral's period of absence. The entry and the mineral returned on the same day, and for a period of some hours the ledger recorded a specimen that the shelf did not hold and the shelf held a specimen that the ledger did not record. The Ordinance now requires all such entries to be made in a stable ink and countersigned after the specimen has been observed. No exception is permitted, including for entries that record the absence itself.

Use in cross-checking

Because an ephemeral mineral's absence is predictable, it serves as a reference standard in Ontological Cross-Checks during Ephemeral Audits: an auditor who cannot find a listed specimen consults the Table to establish whether the specimen is missing or merely away. This has proved reliable in all but one respect. Where a specimen is genuinely lost during its present phase, the loss is indistinguishable, for the duration of the next scheduled absence, from ordinary non-existence, and the audit must be suspended until the mineral fails to return.

The Ledger of Material Cohesion records these suspensions in a hand that varies by auditor. Its binding is a substrate sharing properties with certain entries in the Table, and it is therefore inadvisable to shelve it beside them.

Scholarly disagreement

The principal disagreement concerns whether a mineral absent for its whole recorded period of observation may be entered in the Table at all. The Committee holds that an entry requires at least one confirmed return, since otherwise the substance cannot be distinguished from a blank line. The Office holds that a scheduled absence is itself evidence, and that a mineral absent without interruption for ninety years, as one candidate has been, is the most ephemeral substance in the Tower and deserves the first place in the Table rather than no place at all.

The candidate remains unlisted. Its shelf remains clear. The Office maintains the shelf, and the Committee maintains that it should not.