It had been evident for some time that an updated catalogue of the Charles S. Peirce Papers was needed, one which would survey the whole Collection, making as widely available as possible a detailed statement of what it contained and answering, so far as possible, the questions scholars raise, including those about the date of manuscripts and their relation to published versions. Indeed the manuscripts and correspondence are so voluminous and unwieldy that it is virtually impossible for anyone to deal with them successfully without benefit of the orientation which a catalogue of the kind envisioned would provide. Moreover, as the prospects of a microfilm edition of the Peirce Papers increased, so did the need for an adequate catalogue, which would reflect an orderly arrangement of the Papers and assist the users of a microfilm edition.
The catalogue which was finally produced is imperfect. It is imperfect because of the frequency of error in what already has been done. More importantly, it is imperfect because of what has not been done; that is, much remains to be done by way of identifying and describing, piecing together scattered fragments, assigning dates to undated manuscripts and letters, and the like. But, imperfect as this catalogue is, it is better than none at all, and all of us who contributed to it recognized that the needs for a comprehensive catalogue now outweighed the advantages of indefinite delay.


As noted in the Preface, the Catalogue is divided into two parts. The first part consists of manuscripts and related material; the second part comprises the correspondence, both Peirce's and the correspondence of others. The organization of the correspondence presented no special problems, but the organization of what may be called the "subject matter" part of the Catalogue was another story, and a brief word concerning the problems encountered and the principle of organization finally adopted is in order.
Of the two alternative ways of organizing a man's papers chronologically and by content neither way, in spite of the obvious advantages of each, was easily adapted to the Peirce Collection. Consider the following problems. If the decision is made to order by chronology, what then does one do with the large quantity of undated papers? (Less than half of the 1,644 catalogue entries are dated and of the dates not supplied by Peirce
himself many are conjectural.) Moreover one would have to expect that some of the material would be cut up rather badly as in those instances where Peirce comments on earlier articles. By virtue of temperament and other needs, Peirce can be described as just as Henry James had been an inveterate "revisionist." His tendency to rework drafts of articles and books left future editors of his manuscripts with the problem of unscram-bling the various drafts, which, in some cases, had been written years apart.
Consider now the problems resulting from a decision to order the manuscripts by content. How does one handle Peirce's many digressions? Even more significant perhaps is the problem inherent in schemes that emphasize content; namely, the risk one runs of either imposing too much order or not enough order. Organization is rarely innocent, and the greater the organization the greater the risk that one's bias or interpretation will get in the way of a clear presentation of what there is. However, if one chooses to "play it safe" by arranging the manuscripts as much as possible according to content, thereby achieving a spectrum of sorts, and only then drawing the lines at the more palpable breaks, the results will tend to be nondescript. Finally, as was pointed out to me, if an index were eventually prepared, it would cancel out the need for ordering by content in the first place.
A compromise between ordering by chronology and by content seemed called for. But what compromise? One answer was provided by Boler who, at one point, submitted a plan to the Harvard Philosophy Depart-ment which seemed perfectly reasonable and promising. His plan in-volved six steps: (X) following Burks's bibliography of Peirce's published works (Collected Papers, Vol. VIII, pp. 260-321), locate and file the man-uscripts for each entry; (2) place alternative drafts (and identifiable fragments) with above; (3) from the remaining unpublished material, file what is alike in content with above; (4) also, some of the remaining material, especially complete drafts and identifiable fragments, may be filed chronologically; (5) whenever possible, arrange what remains according to content; (6) finally, classify the remainder of unidentifiable fragments as such. Boler confessed that he became disillusioned about the idea that Steps 3 and 4 would take care of the bulk of the material. I too became disillusioned, and for the reasons Boler gave. But my difficulties with Boler's plan carried somewhat further.
Perhaps the decisive factor in the decision which was ultimately made to compromise while emphasizing content was the fact that the bulk of Peirce's philosophical and other manuscripts the "Archives" material had already been classified by content, in accordance with a scheme adopted by McMahan. The "Houghton" material which had been cata-logued independently by Boler on the basis of some other scheme was from the point of view of both quantity and quality far less significant.
It was tempting, therefore, to adopt the McMahan catalogue, with its principle of organization, incorporating the "Houghton" material as best one could. In this way, the manuscripts might be consolidated, but even more important, since consolidation might be achieved in other ways, was the amount of time and work that could be saved.
The decision to adopt Peirce's own classification of the sciences (which in effect, is what McMahan did) was clearly a practical one, but only in part. Independently there are good reasons for turning to Peirce's classi-ficatory scheme. For one thing, it has the advantage of spreading out Peirce's manuscripts in an orderly way without making the results appear nondescript and without imposing more order than is absolutely necessary. For another thing, it is Peirce's scheme, not someone else's, concocted for the occasion.
There are a number of accounts of Peirce's classificatory scheme of the sciences. In brief, his classification begins with the distinction between a theoretical and a practical science, a distinction based upon the difference of two interests the theoretical interest in attaining knowledge for its own sake and the practical interest in attaining knowledge for the sake of something else. The theoretical branch of science is subdivided into (a) the sciences of discovery and (b) the sciences of review, with the latter dependent upon the former, since review implies the review of something which, in this case, is the information provided by the various sciences of discovery. Indeed, Peirce's own studies in classification are subsumed under (b), as one might expect.
Although Peirce did classify the practical sciences, he was chiefly con-cerned with the theoretical ones, especially those which fell under the heading "sciences of discovery" or, in other places, "sciences of research," and it is his classificatory scheme for those sciences which turned out to be most useful for our purposes. Below is one of several tabular listings from Peirce's papers.*

Phenomenology, or Ideoscopy
Normative Science
Speculative Grammar
Nomological Physics
Classificatory Physics
Descriptive Physics
Nomological Psychics [Psychology]
Classificatory Psychics [Ethnology]
Descriptive Psychics [History]

* This particular list is taken from a manuscript placed with the Matthew Mattoon Curtis correspondence (L107). The manuscript is an incomplete draft of a philosophical autobiography prepared in response to Curtis's request for information concerning Peirce's logical and philosophical views. For a more complete account of Peirce's classificatory scheme for the sciences, see Collected Papers, Vol. I, pp. 75-137. For a good summary account, see Thomas Goudge, The Thought of C. S. Peirce (Toronto: Toronto University Press, 1950) pp. 44-50.

The above listing is for the sciences of discovery (research) only. It should also be clear that the listing is incomplete, for it fails to give the subdivisions of mathematics, metaphysics, and the idioscopic sciences, especially the last with its elaborate arrangement of suborders, families, and subfamilies.
The listing also fails to indicate the hierarchical character of Peirce's classificatory scheme. For Peirce, the sciences listed first are independent of those listed later. Or, if you like, when borrowing occurs, each science tends to borrow from those sciences which precede it in the classification. Thus, for example, in the case of the subdivisions of logic, methodeutic rests upon both critic and speculative grammar, critic upon speculative grammar alone vis a vis the divisions of logic, and speculative grammar upon neither, but only upon those sciences (ethics, esthetics, phenomenology, mathematics) which precede it in the hierarchy. Or, more generally, the mathematician, as such, working independently of the other scientists, seeking formal, not material, truth, traces out the necessary consequences of hypotheses which others, to be sure, may posit. Philosophy (all branches) is dependent upon mathematics, but takes precedence over all the special sciences, which follow it in the hierarchical scheme.
If one examines my table of contents, and observes the order in which Peirce's papers are catalogued, one will note the Catalogue's general adherence to Peirce's classificatory scheme. The Catalogue lists Peirce's mathematical works first, and attempts to deal with these works along the lines suggested by Peirce's division of mathematics into the mathe-matics of logic, of discrete series, of continua and pseudo-continua. The items listed toward the end textbooks, recreations, computations and fragments are conveniently placed there, and have nothing to do with the classificatory scheme for mathematics.
If one ignores pragmatism the next major division of the manuscripts following mathematics and concentrates on the other divisions (phe-nomenology, logic, metaphysics, physics, chemistry, astronomy, geodesy, psychology, linguistics, history, sciences of review, practical science), especially the order in which they occur in the Catalogue, one ought to observe that the remainder of the Catalogue follows Peirce's classificatory scheme, although this may not be self-evident with respect to some of the divisions Why, for example, does chemistry precede astronomy, both in Peirce's scheme and in my catalogue? The reason is that chemistry falls under classificatory physics whereas astronomy falls under descriptive physics, and classificatory physics takes precedence over descriptive physics in Peirce's scheme. Again: Why does linguistics take precedence over history? The answer is that linguistics falls under classificatory psychics, and history, as already indicated, falls under descriptive psychics. Since classificatory psychics precedes descriptive psychics in Peirce's account, linguistics takes precedence over history.
This is not to say that I have slavishly followed Peirce's scheme for the classification of the sciences. As a matter of fact, a rigid adherence to Peirce's scheme is neither required nor desirable. I have followed the scheme only so far as it proved to be advantageous to do so; I have de-parted from it whenever I concluded that by adhering to it the presen-tation of the Peirce material would be hampered Indeed, if one observes closely the organization of this catalogue, one will observe the many liberties taken with Peirce's classificatory scheme, with perhaps the major liberty taken with respect to the manuscripts on pragmatism.
Pragmatism, as a division or heading, presents a special problem. As things stand, given Peirce's classificatory scheme, the manuscripts on pragmatism are out of order. They ought to be in closer proximity than they are now to the logical manuscripts. Pragmatism clearly cuts across the divisions of logic, and perhaps ought to have been subsumed under logic, that is, under one or more of its divisions. After all, did not Peirce come to the view that pragmatism is the logic of abduction? The justification for its present position in the Catalogue, as a separate division between mathematics and phenomenology, rests on the desire not to bury pragmatism among the manuscripts on logic, because of the general im-portance of pragmatism in Peirce's thought and of the lecture series or series of articles of which many of the manuscripts form an integral part.
There are other kinds of problems. One kind concerns the gaps in the Catalogue. To cite one example, Peirce's classificatory scheme calls for the ethnology of social development, one of the sciences comprising one of the many subdivisions of psychical science. The fact that there is no place or listing for it in the Catalogue means simply that none of the manuscripts of Peirce are concerned specifically with the ethnology of social development.
More serious, perhaps, is the failure of this catalogue to provide separate listings for, say, ethics or speculative grammar. But here the problem was not one of finding manuscripts which dealt specifically with ethical problems or the issues of speculative grammar. Indeed there are many such manuscripts. The problem was frequently that of separating units of larger works lecture series or series of articles or chapters in a proposed book something which this editor was reluctant to do. In such cases, the descriptions attached to catalogue entries and the general index are counted on to direct the reader's attention to subject matter for which the Catalogue provides no separate heading or listing.
Then there is the other kind of problem one runs into when dealing with classificatory schemes generally the problem of how to classify this or that relative to the scheme with which one is working. For example, does this manuscript fall under logic or mathematics? Does that manuscript belong with the manuscripts on pragmatism or somewhere else? Often it is not a simple matter to decide, especially when Peirce digresses and when the digression becomes the most significant feature of the manuscript. Sometimes, usually in the case of notebooks, two quite different articles are begun, which forces the editor to decide their relative importance, with the ever present possibility of judgmental error. When confronted with problems of this kind, I have again counted on my descriptions to call attention to anomalies and the general index to bring similar but widely separated material together.
Finally, there are the outright mistakes. One of these will serve as an example. There is no excuse for separating MSS. 314 and 316, since MS. 316 continues MS. 314. In this case the error was discovered only after the microfilming of the manuscripts was completed. Undoubtedly there are errors of this and other sorts which have yet to be discovered. Work on the Catalogue proceeded on the expectation that errors, both of commission and omission, would be made; it also proceeded in the hope that these errors, when discovered, would be reported and collected, and then, in one way or another, made available to users of this catalogue.


The manuscript portion of the Catalogue differs from the correspondence portion with respect to the form employed in presenting the relevant information concerning each entry. For the manuscript portion, each entry is presented in an arrangement of six or seven parts:

1. Title
2. Abbreviated title (Mark)
3. Type of material, whether manuscript, typescript, reprint, or other
4. Publication
5. Date
6. Pagination
7. Description of content

In the Catalogue, Parts 1 and 2 (title) are separated from Parts 3-6 (physical description) which in turn are separated from Part 7 (description of content).
Peirce's titles are presented without brackets or parentheses, just as they appear in the manuscripts. Title page punctuation is retained and the original spellings have been preserved in all titles without the use of sic to indicate deviations from the norm.
The use of brackets indicates that the title has been supplied by the editor. It goes without saying that when a title has been supplied, it is always in the absence of one provided by Peirce, either because he never provided one or because the title page is missing. In defense of supplying titles may I say that it serves as a convenient way of noting a manuscript's principal content and, in many cases, the supplied title as a brief description of the contents saves space by enabling us to dispense with a formal description at the end. May I also add that the supplied titles are sometimes less misleading than the titles which Peirce himself gives. Although Peirce's titles no doubt acquaint us with his intentions, do they also acquaint us with the manuscript's contents? Certainly not in those cases where the manuscript progresses only a few pages and where Peirce's introductory reflections have little or nothing to do with the title. Or, where the manuscript digresses from the topic indicated by the title, and the digression is the manuscript's distinctive feature.
A large number of Peirce's manuscripts have no title, but some of these possess a mark which is most often found in the upper left-hand corner of the manuscript page. When the mark occurs in conjunction with a title, it frequently stands for a short or abbreviated form of the title. It becomes a matter for conjecture when there is a mark but no title. In any event the occurrence of a mark is indicated by the use of parentheses. When the manuscript possesses both a title and a mark, the procedure is to record the title first and the mark in parentheses second. When the manuscript possesses only the mark, then the mark, distin-guished from the title by the use of parentheses, serves in place of the title.
In the next parts (3-6) I was concerned with identifying the type of material, whether a manuscript or typescript, or reprint, or book, or page proof, or galley proof, or the like. I was also concerned with whether, in the case of typescripts, reprints, books, and proofs, there was any annotation or correction.
Most of the manuscripts were not published. But where publication had occurred this is noted by reference to Burks's bibliography and Fisch's two supplements. For an explanation of both Burks's and Fisch's manner of handling bibliographical references, see my explanations of conventions on p. xxvii f. The Catalogue notes whether a manuscript was published in full or in part, and where publication was in part only, precisely what part was published. The only exception to notification of publication occurs in those cases where a part, or even the whole of a manuscript, was published as part of another author's publication. For example, MS. 620 was published as an appendix to one of Fisch's articles on Peirce,* but there is no indication of this publication in the description of MS. 620. This happens to be a significant publication, but, in other cases, it was difficult to say what was and was not significant, and it did not seem worthwhile to mention every publication of this kind.
When not placed within brackets or qualified in any other way, the given date is Peirce's. As a rule one date is given and this is the date which is usually recorded on the title page or, in the case of some note-books, on the cover. Most often it is the only date. But where several dates are given, the range of dates is noted in the description.
When the date is placed in brackets, then the date, as in the case of titles, has been supplied by someone other than Peirce. Whereas I supplied the titles, various persons at different times and with varying degrees of confidence supplied the dates. When the date is placed in brackets without any other qualifying mark, then it is presumed to be accurate, derived from reliable internal evidence. A date preceded by "c." is presumed to be an accurate central locus of possible dates. A date followed by a question mark is frankly a "best guess," based on some internal evidence. When the expression "n.d." occurs, it means that for the moment not even a good guess can be made.
The pagination of a manuscript is indicated by two forms, for example, either pp. 1-5 or 5 pp. The first form signifies that the manuscript was numbered by Peirce; the second form gives the editor's count. One difficulty in determining a true page count rests with Peirce's habit of using the verso of a page of manuscript for calculations or other notes which may or may not be related to the manuscript in question. The question of whether to count a page or not sometimes proved difficult and left room for judgmental error. For additional information concerning pagination, see the guide to the use and consultation of the microfilm edition of the Peirce Papers, prepared by the Harvard University Microreproduction Service, which is reproduced in the next section of this introduction.

* See Studies in the Philosophy of Charles Sanders Peirce, Second Series, edited by Moore and Robin, University of Massachusetts Press, 1964, pp. 24-29.

In 1915, a few of the manuscripts had become separated from the main Peirce Collection. These were added to the general manuscript collection of the Harvard University Library. They were catalogued separately, each with its own call number. Now that they have been restored to the Peirce Collection, their old call numbers have been added to the description for the purpose of identifying them.
In the interest of economy the content descriptions (Part 7) have been pared down to the bare essentials necessary for a clear indication of what there is. The descriptions tend to be topical rather than critical, serving more the function of an index than an analytical table of contents. Not all entries have descriptions, although bracketed titles are intended in all cases to emphasize the principal content of the manuscript. For the most part Peirce's own titles serve the same function. When they do not, a formal description is indicated and provided. But, in general, descriptions are provided for the important entries only, except where the lack of a description means either that, in the case of a draft of a complete or more refined version, the manuscript in question says nothing not already contained in the description of that later or refined version or contains no additional information which in the judgment of the editor is worth special notice. In any event the reader should take note of the number of pages of manuscript. If they are few, the topic or topics indicated by the title or by the formal description may not be very well developed.
Throughout the manuscript portion of the Catalogue, although occurring infrequently, are entry numbers for which there are no manuscripts, as distinct from those entries where a manuscript exists but is missing. These "holes" were created by the fast that the manuscripts which were originally there have been recombined with other manuscripts and that this was done after the completion of the microfilming. Rather than renumber, the entry numbers were retained, but left blank. The "holes" may even have a use someday. They might conveniently serve as the means of slipping new Peirce material into the collection, if such material is ever uncovered.
The correspondence constitutes the last portion of the Catalogue and is divided into four parts: the Charles S. Peirce correspondence, which contains all of Peirce's letters, both those he wrote and those he received; the Juliette Peirce correspondence, which contains all of Juliette Peirce's correspondence, except such correspondence as involves Peirce jointly and which was, for this reason, placed with his correspondence; the family correspondence, which consists of correspondence among members of Peirce's family but which does not involve Peirce or his wife Juliette directly; and miscellaneous correspondence.
The form adopted for the correspondence is the simplest possible one. For the Charles S. Peirce correspondence, the correspondents are listed alphabetically, the number of letters and letter drafts noted, and, when these are dated, the dates recorded, except when more than three of them are involved and when more than three are dated, in which case only the first and last dates are given. Where dates were lacking, an attempt was made to supply them, the procedure here being the same as for the manuscripts. Supplied dates appear in brackets, with or without "c." and with or without question marks. The remaining parts of the correspondence follow the form of the first part.
The division of the Catalogue into two parts manuscripts (or, as sometimes represented, subject matter) and correspondence is a bit misleading insofar as it suggests that no correspondence is to be found in the first part and nothing which is classifiable as subject matter is to be found in the second part. On the contrary, an occasional letter draft may be found among the manuscripts; these were filmed with the manuscripts and all but those which appear on the versos of manuscript pages were subsequently placed with the correspondence, once it became clear that they belonged there. Not all of Peirce's correspondence is personal and business correspondence. There is much which can be described as professional, so much so that if the first few pages and the last were set aside, the remainder could easily be mistaken for manuscript material. Indeed, this is the principal reason why some correspondence was originally placed with the manuscripts.
Finally, a word about the four appendices. Appendix I is a supplement to my catalogue descriptions necessitated by certain discrepancies between the descriptions and what is contained in the microfilm edition of the Peirce Papers. (See the following section of this introduction for an explanation of the discrepancies and the manner of handling them.) Appendix II is a chronological listing of Peirce's manuscripts. It is hoped that this listing can be expanded some day, as scholars are able to date more of Peirce's manuscripts. Appendices III and IV are cross-reference tables. Appendix III is a cross-reference table from Burks's bibliography to my catalogue entries and Appendix IV, from McMahan's catalogue to mine. Anyone who so desires can set out from the Collected Papers and reach my catalogue entries through the intermediary of Burks's bibliog-raphy. See Burks's cross-reference index, pp. 325-330 of Vol. VIII of the Collected Papers.


Two Peirce projects cataloguing and microfilming were linked almost from the beginning. The need for a new catalogue was evident; but so was the need to microfilm Peirce's manuscripts and correspondence, for the physical condition of Peirce's papers was a matter of grave concern. Although the entire collection is now kept in the Houghton Library, where temperature and air control give the papers the best chance for survival, it was feared that even with slightly more handling, given normal wear and tear, the deterioration of the papers would be rapid and alarming. With interest in Peirce mounting and with the expectation that the demand for consulting his papers would most likely increase in the years ahead, it was urged that steps be taken to microfilm them, or at least as much of them as there were funds for.
The success of the microfilming project depended in part on achieving a new arrangement of the Peirce Papers, one which would incorporate the efforts of the past, but would yield a single numerical sequence. With the present catalogue, the numbered sequence was achieved. This permitted the microfilming of Peirce's manuscripts, with all of its advantages of preserving the original manuscript collection from the wear and tear of handling, of providing a record which might serve in place of any parts of the collection that might from time to time be lost, stolen, or destroyed, and finally of making the manuscripts readily available to scholars in all parts of the world.
There are some discrepancies between what was microfilmed and my catalogue descriptions. These are few considering the number of catalogue entries and the principal reason that there are any at all is that errors were discovered in the Catalogue before it was printed but only after the microfilming of the manuscripts was completed. Apart from a major change or two and some minor ones, the microfilm was left un-touched, mainly because of the expense involved in any extensive alteration. An asterisk placed before the catalogue entry number of the manuscript indicates that a discrepancy exists and directs attention to Appendix I "A Supplement to the Catalogue Descriptions."
A short guide to the use and consultation of the microfilm edition was prepared by the Harvard University Library Microreproduction Service in the Fall of 1964. For the benefit of those who will be working with the film and for the additional information concerning the manuscripts themselves, I reproduce the guide here.

This microfilm possesses some apparently anomalous features with which the reader ought to be acquainted to facilitate its use. The major part of the film's unusual features originates in the author's manner of composition.
First it was the author's usual practice to write on one side only of the paper. Less than 5% of the material in this microfilm contained writing on the verso of the page. In the notebooks, Peirce usually wrote only on the recto pages; accordingly, to spare unnecessary expense, only those pages of the notebooks actually bearing text have been filmed. This accounts for the fact that notebooks appear to have been filmed in irregular fashion, sometimes as a single spread and sometimes as a double spread. A similar situation prevails with the material written on loose sheets. In a few instances, both with the notebooks and the loose sheets, Peirce used the opposite sides to make routine calculations, some related and some unrelated to the main body of the work. In most instances, these routine calculations have not been filmed. Where there was doubt about routineness or where the calculations were other than ordinary arithmetic, such material was microfilmed. Some of these data may thus appear to interrupt the normal sequence of the manuscript.
Another unusual feature concerns pagination. The manuscripts fol-low four schemes of pagination: (X) unpaged, (2) either even-numbered or odd-numbered, (3) normal, and (4) iterated pagination. The re-peated pagination almost always occurs in the notebooks when Peirce was constructing a draft If he was dissatisfied with his first draft of page 1, he would go on to the next page, number it also "page 1,'' and continue with his revision until satisfied that he could carry on with page 2, and so on It is not uncommon for a page number to be thus repeated for four or five consecutive drafts before the next sequential number.
Odd-numbered pagination only is common in the notebooks. Evi-dently this was Peirce's way of indicating his consciousness that he was using only the rectos, or perhaps he was saving the versos for cor-rections or changes. In a few instances, an explanatory target accom-panies each frame of film and states that no pages are missing.
Unpaged material has been placed in sequence insofar as this was ascertainable by the editors, and, of course, insofar as the actual pages were available.
At the end of a numbered sequence of pages, there will occasionally be found a miscellany of pages consisting of broken runs or isolated pages surviving from other drafts.
Another unusual condition arises from Peirce's practice of starting some notebooks from the front, and upon reaching the center, turning the notebook upside down and beginning anew from the "back." Sometimes the separate contents of such notebooks may be unrelated although they occupy the same physical and bibliographic unit; in other instances, after the notebook was turned upside down, the same material was continued. This condition prevails in little used as well as in full notebooks. Rather than inconvenience the reader of the film with upside down images or reversed pages sequences, all such material has been filmed for normal reading sequence. In each case a notice explaining this situation is filmed at the beginning, the center, and the "end" of the item.
Peirce occasionally constructed from paper a physical device to be removed from a notebook and manipulated. An example is a dough-nut-like device he constructed to elucidate a point in topology. In filming devices, a first exposure has been made with the device in place, a second with the device removed, and if necessary for clarity, a third of the device itself.
Printed editorial forms used in connection with the partial publi-cation of this material by the Harvard University Press in the Col-lected Papers have remained with the collection, and it is possible that a few of these may have been accidentally incorporated into the micro-film. These are of course not a part of the collection and should be ignored.


Generally speaking, a catalogue of a man's writing stands as an impersonal record of his achievement. Standing alone it seems to cry out for some kind of personal statement, a portrait of sorts, which would complement the impersonal record. Of course it is a matter of conjecture as to what kind of personal statements or portrait of himself Peirce would have appreciated. In the introduction to a catalogue a panegyric seems somehow out of place. Perhaps it would be best to let the catalogue speak for itself. The display of prodigious intellectuality, creative genius, philosophic and scientific integrity, demonstrated therein, and, for one who knows something of the frustrations and deprivations of Peirce's personal and professional life, the sense of tragedy that pervades the whole seem to me to be intellectually stimulating and, at times, profoundly moving.