How to Read an Organbuilder Opus List

An opus list records the numbers a firm assigned, in the order it assigned them, and nothing in that definition promises that each number became a finished organ that still exists. The published Skinner, Aeolian-Skinner, and Ernest M. Skinner & Son lists make this explicit: they carry numbers for instruments that were not built, numbers for which no record survives, numbers whose contracts were transferred between corporate entities, and suffixed numbers recording later work at a location already on the list, and in the archived Skinner and Aeolian-Skinner lists the years shown are contract years rather than completion dates. Those conventions are stated for those particular lists; another builder’s list may follow different rules and has to be read on its own terms.

The same character shows in a live builder list. As of August 31, 2026, Austin Organs publishes a searchable opus list on its own website giving opus number, state or country, city, venue, number of manuals, and number of stops, and it includes entries flagged “Organ not built” or “Contract Cancelled”. A register that keeps its cancelled contracts is doing what an opus list does: it preserves the sequence of numbers, and it leaves the question of what was actually built and what remains to other evidence.

A single entry establishes that the builder issued that number against a named place, with whatever descriptive fields the list carries, and that is the whole of its direct testimony. The fields differ from list to list. Austin’s table gives venue, manuals, and stops; the Orgues Létourneau list gives client, location, country, ranks, manuals, and action across seven columns; the public Estey list gives opus number, year, city, state, and original location in five. Where the list gives a year, read the list’s own statement of what that year is. The archived Skinner and Aeolian-Skinner lists say plainly that the years shown are the years contracts were signed and do not reflect years of installation, which is why a listed year can differ from the year on a dedication plaque without either being wrong. What each date type does and does not establish is the subject of Contract Date Versus Completion Date.

Some lists give no year at all. Across both pages of the Létourneau list on September 2, 2026, none of its 142 rows carried a build year, a completion year, a dedication year, or any other date, and neither of the firm’s two published instrument indexes, its opus list and its opus map, supplies a date. An entry on a list of that kind does not date the instrument, and it does not even establish that the instrument is finished: the highest-numbered Létourneau entry on that date, Opus 143, is scheduled by its client for installation in 2027. An entry is a number issued against a contract; completion, installation, and survival are separate facts that the list may or may not record.

Nor does an entry settle specification. The maintainer of the Estey list notes that the shop orders written at construction time generally contain more complete information than the opus list, and that in some cases the two records vary, so a single list entry is a pointer toward the record, not a substitute for it.

A gap in an opus list means that no entry was published for that number, and documented reasons include an instrument not built or a number for which no record survives. The archived Skinner and Aeolian-Skinner lists mark some numbers NB, for an instrument not built, and NR, where no record exists, so a coded or empty number does not by itself mean an instrument is missing or destroyed. Austin’s live list keeps its unbuilt and cancelled contracts in place under their own numbers, which has the same effect: the sequence is preserved and the reader is told why a number produced no organ.

Some gaps carry no code and no explanation. On September 2, 2026, the Létourneau list ran from Opus 104 directly to Opus 102 with no row for Opus 103, and the firm publishes no statement about gaps in its series, so no reason for that omission is established. A gap of that kind is an open question, not a finding. Read it as a number about which the published list says nothing, and treat whatever instrument may stand behind it as undocumented by this list rather than as nonexistent.

Two numbers against one project are a documented pattern in at least one major builder’s own records, and the rule behind the second number is not stated. The Casavant Frères history timeline lists Opus 1587-1588 for a 1938 installation in Lewiston, Maine, and Opus 3812/3813 for a 2001 installation at Central Synagogue in New York, and a rebuilder’s project record for Grace Episcopal Church describes a 1922 Casavant acquired as Opus 940 and 941. None of those sources explains what the second number denotes, whether a separate console, an antiphonal or echo division, or something else, so the pairing is an observable feature of the list rather than a documented convention.

A pair on one project is different from two numbers in one building. Compiled New York chapter documentation for St. Patrick’s Cathedral gives the gallery organ as Kilgen Opus 5918 of 1930 and the chancel organ as Kilgen Opus 3920 of 1928: two contracts, two instruments, two numbers, in a building that a casual search would return once. An opus number identifies a contract, not a site, and a building with more than one instrument can legitimately carry more than one number.

Later work can return a number to a list rather than add a second new organ. In the Skinner and Aeolian-Skinner lists a suffixed opus number records later work at a location already on the list, keeping the base number of the earlier instrument. Kilgen went further and assigned opus numbers to supplementary orders: a theater-organ society record for the Asbury Park Kilgen, Opus 4258 of 1929, lists further parts, a generator, swell shades, and a relay rebuild supplied under Opus 4610, 4613, 4619, 4646, and 4717. Five numbers there stand for parts and repairs on one instrument, which is one reason a count of numbers issued is not a count of organs.

A rebuild can also receive a fresh number in the rebuilding firm’s series. The 1942 Kilgen Organ Company rebuild at Lafayette Park United Methodist Church in St. Louis is recorded as Kilgen Opus No. 7050 on an instrument originally built by Geo. Kilgen and Sons in 1904, so the list carries two Kilgen numbers for what is physically one organ with two layers of work. How to read a rebuilder’s number, a secondary plate, and the layered identity they produce is covered in Rebuild Plates, Secondary Opus Numbers, and Layered Instrument Identities.

Published opus and year pairs do not reliably run in number order for every builder, and the documented exceptions are unresolved observations rather than evidence of a numbering policy. Published Kilgen pairs include Opus 3004 (1919), Opus 4049 (1928), Opus 4258 (1929), Opus 5163 (1933), Opus 5395 (1935), Opus 5840 (1927), Opus 5913 (1938), and Opus 7722 (1955). Opus 5840 carries a year eight years earlier than the lower-numbered Opus 5395. Each pair comes from a different secondary source: the 1927 year for Opus 5840 is from a third-party database record, and the 1935 year for Opus 5395 is from a 2023 academic document on that instrument. The inversion may reflect Kilgen practice or an error in one of the separate sources, and neither explanation is settled, because no Kilgen company document setting out a numbering rule is available to decide between them. The conclusion a reader can safely draw is narrower: a Kilgen opus number on its own does not date the instrument without checking the list and the record behind it.

The pattern is not unique to one firm. Third-party listings of Schantz opus and year pairs contain non-monotonic and apparently anomalous examples and are research leads, not a dating rule. A regional survey of Wicks instruments reports five opus and year pairs from 1973 to 1994 and one internally anomalous pairing, Opus 3447 with 1980, which should be treated as unresolved.

One documented cause of an anomalous pair is a conflict inside a single database record. A Pipe Organ Database record for First Presbyterian Church, Morristown, Tennessee, read through an unaffiliated mirror, carries a header of Schantz Organ Co., Opus 1116, 1973, while the stoplist attached to the same record is labeled Op. 1306 and 1974. The indexed field and the document it was built from disagree, and the record does not say which is right. That one case shows the mechanism; it does not say how often it happens. When an opus and year pair looks wrong, check whether the number and the year came from the same document before concluding that the builder numbered out of sequence.

A list’s first number and its dividing lines are editorial choices of the compilation, not necessarily events in the firm’s history. The published Skinner Organ Company list runs from Opus 101, Unitarian Church, Ludlow, Vermont, contract year 1902, to Opus 872, Girard College Chapel, Philadelphia, contract year 1931, so the Skinner series does not begin at Opus 1. The archived Aeolian-Skinner list picks up at Opus 873, but Opus 873 through Opus 878 carry the contract year 1931 and only Opus 879 and above carry 1932, so the earliest instruments on the Aeolian-Skinner list were contracted before the Aeolian-Skinner Organ Company, Inc. existed. The break between the two lists is a convention of the compilation rather than the date of the corporate change.

The same Skinner series also runs under several corporate names, so an opus number in that range does not by itself say which entity signed the contract. Which corporate identity a number belongs to, and why two firms can hold the same number, is the subject of Why Two Builders Can Share the Same Opus Number.

A milestone timeline is a selection, and the continuity it suggests between the milestones is implied rather than demonstrated. The Casavant timeline presents a single continuous opus series running from Opus 1 in 1880 to Opus 3905 in 2013, and a trade news item describes Opus 3920 at a concert hall in Fuzhou, China. Nothing in the timeline is wrong as far as it goes, but it is a chosen set of landmarks, not a complete opus list, and it cannot be used to fill in the numbers between them.

Milestones of output are a different quantity again. The Dictionary of Canadian Biography records Casavant production peaking at 55 organs manufactured in 1929, and the Encyclopedia of Music in Canada gives cumulative output milestones of 100 instruments by 1899, 200 by 1904, 500 by 1912, 600 by 1914, 1,000 by 1923, and 1,355 by 1929. Cumulative output is not the same thing as the opus number reached, and neither reference work says the two coincide. A reader who lines up an output milestone against an opus number from the timeline is joining two series that no source has joined.

A published list can cover only part of the numbering range the firm itself reports, and the shortfall can be large. As read on August 31, 2026, the scope of Austin’s live opus list ran from approximately Opus 1 to Opus 1629, against roughly 2,800 instruments the company reports having built. On September 1, 2026, the public table still reached Opus 1629, while The Diapason has reported the original company’s final contract as number 1885 and a later Austin instrument as Opus 2790. An Austin whose number lies above the range the public table showed is outside what the table covered on that date, which says nothing about whether the instrument exists.

Instruments can also exist outside the numbered series altogether. Jonathan Ambrosino, writing in 1993, states that M. P. Möller had built more than 12,000 organs by the end of the firm, that the last was numbered 11,860, and that several hundred were built outside the opus cycle. That is one identified practitioner’s account rather than a universal rule, but it establishes that an unnumbered or off-cycle Möller can exist, so the absence of an opus number does not by itself disqualify an instrument from being a Möller.

The highest number a list reaches, the count a firm publishes, and a cumulative output figure are separately measured quantities, and none of them converts into the others. Published counts for Orgues Létourneau show the point: James B. Hartman recorded 55 organs built to 1998 with 13 more in progress; The Diapason reported 132 organs completed at the November 2019 ownership announcement and, separately, that the founder had built over 140 new pipe organs; the APOBA member listing says over 130; and the firm’s own opus list reached Opus 143 as read on September 2, 2026. These figures are not contradictory once their dates are respected, but they are not interchangeable and should not be averaged or quoted without their date.

The same list shows why a row count is not a count of finished organs. On September 2, 2026, the Létourneau list carried 142 rows against a highest number of 143, with Opus 103 absent, while the same page introduced itself with the firm’s own sentence, “More than 140 new Létourneau pipe organs have been designed and built in our workshops,” a claim about instruments designed and built rather than a count of rows. The 142 rows, the firm’s “more than 140,” and the APOBA directory’s “over 130” on a page last modified September 8, 2023, are three separately made statements and should be kept apart rather than reconciled. And since the highest row, Opus 143, belongs to an instrument its client schedules for installation in 2027, at least one row is an instrument not yet built.

Möller’s lifetime totals make the same point at a larger scale, and each figure counts a different thing. Ambrosino’s 1993 account gives more than 12,000 organs with the last numbered 11,860 and several hundred outside the opus cycle. The Organ Historical Society’s finding aid for the Möller collection states approximately 12,000 pipe organs. OHS Press describes Bynum Petty’s 2023 company history as carrying a geographical index of 11,850 organs. As of September 1, 2026, the OHS Pipe Organ Database held 13,118 instrument entries for M. P. Möller, a figure that includes separate entries for rebuilds and additions on the same organ and that changes as records are submitted. Organs built, organs indexed, the last number issued, and database entries are four measurements, not four estimates of one number, and they are not to be reconciled into a range.

Before relying on a list, establish what it was derived from, who maintains it, and when it was last revised, because a frozen derivative and a maintained register look identical on the screen. The public Estey opus list states that it derives from the list E. A. Boadway prepared from factory cost cards for publication by the Boston Organ Club in the 1970s, with updates since; its maintainer holds photocopies of Estey shop orders for Opus 760-900, 1004-1356, and 2051-3261 with a few exceptions; and the page states it was last updated June 8, 2015. That is a derivative of a 1970s compilation, backed for three declared opus bands by records in private hands, and unchanged for more than a decade.

The same list circulates in more than one version. The Estey Organ Museum publishes the list as a 57-page PDF under a 2020 upload path and a filename carrying the revision date 3-14-2020, describing it as compiled by Boadway and provided by the Organ Historical Society. The 2020 date comes from the filename and path, not from a revision statement inside the document, which has no changelog, and a later distribution date does not prove later content; where the two versions have been compared at individual entries they agree, Opus 89 reading 1903 in both. The list was also corrected in print after its first publication: the May to December 1977 Boston Organ Club Newsletter amended entries from the January and February 1974 installment, correcting Nos. 108 and 109 and updating the whereabouts of the pipes of No. 1318 and the home of No. 2225. A reader holding one installment alone would not see corrections printed three years later, and a correction printed in the newsletter cannot be assumed to appear in the consolidated PDF.

Some lists declare their own incompleteness. The Robert Morton opus list says its compiler intends to establish a list for the entire output of the firm, by state, and warns that entry is by hand and will take a while, so absence from that list may mean an instrument is undocumented or simply that its state has not yet been keyed in. That wording was current as read on September 2, 2026, but dates from no later than the page’s 2019 update.

A database is a different custodian again. As of August 31, 2026, the OHS Pipe Organ Database records Möller instruments in the form “M. P. Möller – Opus NNNN, YYYY” with institution and location and hosts Möller factory documents, such as an “Original Specifications” sheet for Opus 8414, but its instrument pages require JavaScript and its legacy detail addresses return HTTP 410. A contributed database entry with attached documents is not the builder’s register.

Versions of one list can differ in the fields they carry, and a field dropped in a later version has to be recovered from the earlier one. The 1973 to 1979 newsletter printing of the Estey list gave the manual count for each entry, and its compiler stated that an “A” following the number of manuals indicated a player attachment; the 2020 museum PDF and the frozen web version present only opus number, year, and location. Why the field was dropped is not established, and a manual count is a size indicator rather than a specification, but a researcher who needs the player marking has to go back to the newsletter scans.

Column sets also differ between builders in ways that limit what an entry can answer. The Létourneau list’s seven columns include ranks, manuals, and action but no date of any kind; Austin’s table includes manuals and stops and a flag for unbuilt or cancelled contracts; the Estey list gives location fields and a year. Match the question you are asking to the fields the version in front of you actually carries.

A live list is read as of a date, and any figure taken from it is a reader’s observation on that date rather than a figure the builder publishes. Even fully traversed, the Létourneau opus list gives no total entry count, no revision date, no completeness statement, and no inclusion criteria; the 142-row count is a count made by a reader on September 2, 2026. The Robert Morton page may since have dropped its in-progress wording or declared its coverage complete, and the highest Létourneau number will move as the firm adds rows.

The practical rule is to record the access date with every observation taken from a live list, to recheck the list before the observation is relied on, and to treat the highest number on it as the highest number published on that date rather than as the firm’s current output.

A mismatch between an instrument and its list entry is a prompt to corroborate, not a verdict against either. Start with the date: where a list states that its years are contract years, a later dedication year on the instrument is an expected difference rather than a contradiction. Then separate layers of work: a builder’s own production count may state instruments built and rebuilt together, and a rebuild can carry its own number in the rebuilder’s series, so the number on the console plate and the number on the list may belong to different events in the same organ’s life.

Treat the surviving instrument as a filtered sample of what the list recorded. The records document instruments being relocated, reduced in size, converted to a different action, rebuilt by other firms, and lost, so the organ in the building may match its entry in number and place and still differ from it in size, action, or contents. An entry with no matching instrument, and an instrument with no matching entry, are both consistent with a list that registers numbers issued rather than survivors.

What still needs corroborating is the link between the number and the object, and the list cannot supply it, because the record behind an entry, the shop order or contract file, can carry more than the entry does and can differ from it, as the maintainer of the Estey list states of that list. That corroboration comes from the plates, stamps, and marks physically present on the console and inside the instrument, which is a matter of where to look and how to inspect; from the contract file, shop order, or specification held in an archive, which is a matter of holdings and access; and from the corresponding database record with its attached documents. Where the instrument holds more than one builder’s work, deciding the attribution and ranking the evidence are their own subjects. A number that matches an entry on a different firm’s list is not a match at all; see Why Two Builders Can Share the Same Opus Number.

When an opus list entry and an instrument disagree, the useful next step is a documented account of what is actually in the building, the plates and marks, the action and windchest types, the console, and the pipework, set beside the list entry and whatever archive record can be obtained, so that the identification rests on the instrument rather than on a row in a table.