The Works

Brieven or Werken
Opera Omnia

There is no Dutch Werken or Latin Opera Omnia in the sense of a set of four volumes uniformly edited, typeset, printed, and bound by the same shop at the same time while Leeuwenhoek was alive. For libraries and collectors, as long as the set has all 165 letters in either Dutch or Latin, it is complete. See the Case Study: Grievous difficulty: shuffling Letters 28 - 52 for a detailed comparison of several sets.

Until Dobell, Harting's 1876 bibliography was the most comprehensive. He provided conventional bibliographic data for the pamphlets and collected volumes. However, when it came to the Werken, Harting noted only two 4-volume sets and two 5-volume sets without specifying their contents or differences. The National Library of Medicine lists a 5-volume set:

• vol. 1: Dobell #8, 2, 5a, 6, 7, 9
• vol. 2: Dobell #10a, 12, 13
• vol. 3: Dobell #14, 15, 16
• vol. 4: Dobell #18
• vol. 5: Dobell #19

It has Dobell #1 inserted in chronological order between the two parts of #8. However, it seems to be missing Letter 39 and Dobell #3 and #4, Letters 40 and 41.

The difference between a four-volume set and a five- or six-volume set seems to be in the middle volumes. If #14 is bound in volume 2, then #18 fits into volume 3 and #19 Send-Brieven is volume 4. The National Library of Medicine also lists a 6-volume set that has the middle volumes in pairs: Dobell #10 and #12 in volume 2, Dobell #13 and #14 in volume 3, and Dobell #15 and #16 in volume 4, pushing Dobell #18 and #19 to volumes 5 and 6.

According to Dobell's bibliography (1932, p. 394), this is #20:

Brieven [seu Werken]. 4°. 4 vols, (or sometimes 5). Various dates, publishers, and places. — The final Dutch collective edition of all L.'s published letters. Contains Letters 28-146 and I-XLVI, and is variously made up of the several separate issues already listed—bound together.

In other words, Werken and Opera Omnia are more accurately thought of as bindings or bundles or collections than as separate publications. As Dobell notes (p. 396), "many other collections are to be found". For example, Dutch scientific instrument collector Bert Degenaar has a set (above; click to enlarge) that he was kind enough to let me examine in person. It was the set that Abraham Schierbeek owned and used to write his two-volume biography in 1950. It differs from Dobell's, as detailed on the table below but not noted in Schierbeeck's own bibliography.

For the Dutch versions, it is thus difficult to assign a single publication date for the Werken, per se. The table below has the dates of first publication of the component volumes in Dobell's set, all of which are on the pages Works Part I and Works Part II.

For the Latin versions, what Dobell rightly considered the Latin equivalent of the four-part Werken is not a direct equivalent, that is, the each volume in both series has the same letters. Instead, previously printed volumes, containing various numbers of letters, were divided into four groups. The Opera Omnia is the complete set of all of the components in Latin.

Parts III and IV of both language series are identical. The variance stems from the only letters that van Gaesbeeck published in 1684, six of them. They were not reprinted in Dutch or printed in Latin at all until the mid-1690s. In Dobell's set, they're in Werken Part I but in Opera Omnia Part II.

As an example of the variation in the bundles, Dobell's copy (1932, p. 396) of Opera Omnia I has sixteen letters (28-31, 34-36, 44-52) printed by Boutesteyn in 1687 even though Langerak had re-printed them in 1722 (Dobell's 23a). See examples of Bundles on the right sidebar.

In the custom of the time, the titles were very long. Printed separately, the title pages served as advertisements for the books. They were hung on walls and passed out at the book fairs.

The titles are shortened and translated for Lens on Leeuwenhoek. Dobell's numbering, the standard, is used throughout.

On the table below, each letter is accounted for once in each language, with the exception of Deel I, on the first row. STCN notes a 1696 second edition of Deel I -- Ontdeckte Onsigtbaar-heeden, published by Boutesteyn alone, not Boutesteyn and van Gaesbeeck. This edition is not in Dobell's or Schierbeek's bibliographies. By then, Boutesteyn had reprinted the six letters that had been printed by Gaesbeeck in 1684, so he could publish his versions with this volume. The STCN notes that this title is held by a number of libraries; some have the letters bound in numerical order (new pagination, too?) and others are incomplete.

The table below shows the variaton in two sets considered to be complete Werken and Opera Omnia. Because Dobell notes letters out of order in other places, I can only assume that his Part I was bound in the numerical order of his volumes (column 2 and title pages on left sidebar) and not the order of the letters. However, his bibliography notes:

Deel I —Made up of Nos. 1, 2, 3, 4, 5, 6, 7, 8, 9 : thus containing Letters 28-52. With engraved title dated 1685 (1st state, with Dutch lettering Ontdeckte Onsigtbaar-heeden).

Degenaar's set, once owned by Abraham Schierbeek, contains the corresponding letters in Part IV. This volume seems to have no order and includes another edition of 10 Vervolg der Brieven that is in his Part II. The version available online at Google Books has the same order as the Degenaar volume on the table below.

Note in the photo above right that the Degenaar volumes all have Send-Brieven on the spine. Indeed, the volume that Leeuwenhoek titled Send-Brieven, containing letters written at the end of his career, is the first part of Degenaar's set and the earliest letters are the final part.