“One certainly never leaves his auditorium without taking back home
many illuminating hints regarding his writings.”
— J. S. Purgstall (1773-1812)
Overview of Kant’s Lecturing Activity [top]
|logic||ooooooooo||oooooNoooooooo ooooo||o Ñ N o o N o o o o||Ñ Ñ Ñ Ñ Ñ o o o o Ñ||o o N N N o o||56|
|metap||ooooooooo||ooooNo Ñoo ooooooo o||ooo o o o o o o Ñ o||o o N o N o o o o Ñ||Ñ Ñ N N o||53|
|geogr||o ooo Ño||oo NN oooooooooo||N oooNo N N o N o o||o o Ñ o Ñ N o o o o||o N N N o o Ñ||49|
|anthr||N o o N o N o o||o N Ñ o N o o o Ñ Ñ||o N Ñ Ñ o o||24|
|physi||oooooo o||o oo N o o o o||N o||o o N o||21|
|moral||o o oo||o o Ñ o ooo o o||o oo o Ñ o oo o||o Ñ N o o||N||28|
|law||o o||o o o o o||o o N o o||12|
|encyc||o oo||oo o o Ñ o||o||10|
|theol||o||N Ñ o||4|
|pedag||o||o o o||4|
|mathe||oooooo oo||oooo Ñoo||15|
The disciplinary links in the left-most column lead to the appropriate section of the List: Kant’s Lectures by Discipline.
The columns are grouped by decade, with summer semesters arranged under the digits (in red) and winter semesters arranged under the dots (in black). For instance, the ‘0’ in the 1770s column refers to the summer semester of 1770 (these data are all presented in red type), while the ‘’ between the ‘0’ and the ‘1’ refers to winter semester 1770/71 (in black).
The table begins with a dot (), under which are the courses Kant taught during his first semester (WS 1755/56).
Lectures (and available notes from that semester) are indicated with either of three symbols — ‘o’, ‘N’, ‘Ñ’ — depending on whether we possess lecture notes from that semester, and the certainty we have of the dating. An ‘N’ indicates that we possess lecture notes for which we are reasonably certain they stem from this semester, while an ‘Ñ’ (or N-tilde) is to be seen as a placeholder for notes that stem from roughly this period. [N.B.: Holding your cursor over the ‘N’ and ‘Ñ’ should cause the names of the relevant notes to appear.] These approximations need to be viewed quite cautiously, as often there is considerable disagreement on the dating. Also, many of the notes are compilations from various semesters (see the page on Dating the Notes); I list these under each relevant semester. Much of this dating information should be seen as provisional, and is likely to change with further research. In any event, the discussion on dating included with the separate descriptions of the notes should always be consulted.
The very bottom row totals the number of courses taught during any given semester (courses were held four hours each week, except for Pedagogy, which was only one or two hours). These totals include an occasional unnamed privatissima (see the caveats, below), and so will sometimes be more than the number of courses displayed in the table. Total number of courses taught, by decade: 42 (1750s), 84 (1760s), 69 (1770s), 61 (1780s), and 26 (1790s).
The far right column adds up the total number of times a course of lectures was taught (Kant lectured 56 times on logic, 53 times on metaphysics, and so on). But see the caveats, below.
The table does not include the Practicals (variously called repetitorium, examinatorium, disputatorium), which typically met twice each week, on Wednesday and Saturday, but in later years only on Saturday. We have only sporadic evidence of these prior to 1770, but Kant is shown to have offered one nearly every semester after that, and we have reason to believe that he always offered these practicals [see Borowski, qtd. in Reicke 1860, 32; repr. Malter 1990, 43].
Caveats. This table is generous with the data, including all courses that likely took place, even though not all of them (especially those prior to 1770) are confirmed; similarly, official data for WS 1758/59 are not available, perhaps as a result of the changes in government with the beginning of the Russian occupation [glossary]. The totals in the far-right column for logic and metaphysics are higher by one more than the tallies of their respective rows (in each case because of a semester where Kant appears to have taught two sections of the same course — 1770 for logic, 1771/72 for metaphysics). Similarly, the tallies in the bottom row include additional, unnamed privatissima for certain semesters (SS 67, SS 69, WS 69/70 [twice], WS 70/71) as well as the doubled sections of logic (SS 70) and metaphysics (WS 71/72). The 279 at the bottom of the far-right column is a simple total of the numbers in that column. If the five unnamed privatissima are included, the total number of courses taught will be 284.
Kant lectured for 41 years, or 82 semesters beginning with the 1755/56 winter semester (WS) and ending in the middle of the 1796 summer semester (SS) and he gave an estimated 284 courses of lectures (not including the weekly practicals), for an average of 3.5 courses per semester. All of this took place in Königsberg, a cosmopolitan city on the Baltic Sea where Kant was born and raised. The university was called the Albertinum [glossary], and it was here he spent his entire student career and from which he eventually received his Magister degree (on 12 June 1755) and where he habilitated (on 27 September 1755), which gave him the right to lecture. [more]
For the first 29 semesters of his teaching career, from WS 1755/56 to WS 1769/70, Kant taught as a lecturer [Privatdozent], offering only private courses where he was paid directly by his students at the end of each semester in the form of 4 rthl. honoraria, since such lecturers received no salary from the government. He was also free to teach what and when he wanted, constrained only by the willingness of the students.
Upon assuming the full professorship of Logic and Metaphysics (beginning with WS 1770/71), Kant was required to offer one public (free) course of lectures [glossary] every semester from 7 to 8 each Monday, Tuesday, Thursday, and Friday (the four principal days for instruction) — specifically, a course on logic each summer semester and a course on metaphysics each winter semester. He was also free to offer private lectures, which he did, although with less frequency than when he taught as a lecturer, and by the late 1770s he was teaching just three courses per semester in a fairly regular pattern, alternating between Logic and Physical Geography in the summer and Metaphysics and Anthropology in winter, and adding a third course on a four-semester cycle, alternating here between Philosophical Encyclopedia, Natural Law, Moral Philosophy, and Physics. In his last seven years he dropped this third class.
Apart from public and private lecturers, one could also offer privatissima lectures [glossary], where a group of individuals would privately contract with the professors to give a course of lectures. Records indicate ten privatissima that Kant held: mathematics (WS 1763/64), philosophical encyclopedia (WS 1768/69 and SS 1770), metaphysics (WS 1771/72), physical geography (WS 1772/73), and then five unnamed courses (1767, 1769, 1769/70 (two courses), and 1770/71). Apart from this, there is also mention of Kant offering, in his early years, privatissima courses on fortification to military officers (see).
Kant was not the only instructor offering lectures on philosophy at Königsberg (see Philosophy Faculty at Königsberg), and it appears that there was considerable competition for students. For instance, in any given semester there might be four or five private courses on metaphysics, in addition to the course being offered free of charge by the full professor likewise for courses on logic, practical philosophy, natural law, natural theology, physics, and so on. Although contemporary accounts speak of Kant filling his classrooms to overflowing, the records show that he often failed to do this. Almost eight percent of his private offerings were cancelled for lack of sufficient student interest; of these cancellations, about two-thirds (fourteen) were replaced with courses that actually took place, while the remaining third (seven) were not.
When considering how many courses Kant was teaching in any given semester, keep in mind that instructors would occasionally fail to finish a course of lectures during one semester and so would carry it over into the next semester. This happened often enough to elicit a censure from the government that instructors were to be more diligent in completing their courses, so that students could finish their coursework on time (i.e., in three years). The records suggest that Kant carried over some of his classes into a second semester only during his earlier years, for instance, during the SS 1766 Kant was finishing up his courses on metaphysics and moral philosophy from the previous semester, as well as beginning new courses of lectures in each of these subjects (this was in addition to lectures on logic and physical geography). His strategy this semester was to finish up the two courses on Wednesdays and Saturdays, so that the main lecture days were available for starting the four new courses.
Most of our information regarding Kant’s lecturing activity comes from Arnoldt and Schöndörffer [Arnoldt 1908-9], and this information is much less reliable for Kant’s fifteen years as a Privatdozent than for the twenty-five years as a professor, primarily because of institutional changes in how records were kept.
Arnoldt worked in the 1890s from records stored at the university in Königsberg and in Berlin at the Geheimes Staatsarchiv. Schöndörffer had access to the same records used by Arnoldt, as well as to records kept in the Budget Ministry (Etats-Ministerium) in Königsberg — the local government office charged with overseeing the university. The kinds of records stored in each of these sites were the same, but each set had various gaps, and so the use of all three rewarded Schöndörffer with additional information. In general, the university generated three sets of all their records, keeping one set and sending the others to Berlin and to the Königsberg Budget Ministry. After World War II, the university archives were believed to be lost, but in the spring of 1990 they were located in the Polish State Archives in Olsztyn (German: Allenstein). The records Arnoldt used in Berlin had been stored in Merseberg, and only recently have been returned to the Geheimes Staatsarchiv in Berlin-Dahlem. Five kinds of records are relevant to Kant’s lecturing activity at the university:
|(1) M. Pisanski, 10-11, Isagogen|
|(2) M. Weymann, Praelectionum Metaphysicam, Logicam, Moralium et Jur. nat. et. gent. 10-11, 4-5 + Wed/Sat 3-4|
|(3) P. Reusch, Metaphysics and Logic, private, 10-11.|
|(4) M. Wlochatius, Logic (using Crusius), 8-9.|
|(5) P. Buck, Metaphysics, private (using Feder), 2-3.||Taught.|
|(6) P. Kant, Metaphysics, public (using Baumgarten), 8-9.||Taught.|
|(7) M. Wlochatius, ibid. (using Crusius), 9-10.||Failed.|
|(8) M. Johswich, ibid., 9-10.||Failed.|
|(9) P. Reusch, Theoretical Physics, public, 3-4.||Taught.|
|(10) P. Kant, ibid., private (using Erxleben), 9-10.||Failed.|
|(11) M. Johswich, ibid., 2-3.||Failed.|
|(12) P. Buck, Experimental Physics, private (using Feder), Wed/Sat, 11-12 and 2-3.||Taught.|
|(13) P. Reusch, ibid., private, 9-10.||Taught.|
|(14) M. Wlochatius, Physics (using Crusius), 10-11.||Failed.|
|From a handwritten record (prepared Sept. 7, 1772 — classes for WS 1772/73 would begin in five weeks) in the Acta at Allenstein (APO XXVIII/1, Nr. 200). These are the courses listed under "d. Philosophical" (the courses listed under "c" appeared to be mathematics courses). Whether a course actually took place comes from a Consess of February 26, 1773. The reason given for a course failing was always "low enrollment."|
Preliminary lists of the courses (lectures and practicals) announced by the professors and lecturers for the coming semester. These lists appear to have been prepared in the few weeks before each new semester. For instance, SS 1766 classes began April 14, and the list of courses was prepared five weeks earlier on March 10. They come in several versions, however: as a list of a professor’s own offerings for that semester written in his own hand; as a composite list, with each professor entering his own courses; as a composite list written in a single hand likely that of a dean (if the courses are just of one faculty), the rector, the philosophy dean (writing on behalf of the rector), or some anonymous copyist.
Vorlesungverzeichnis is the official lecture catalog for the semester, printed in Latin and based on the above preliminary lists. Current versions were distributed to newly matriculating students, and the new catalogs would be printed and posted on the university message board no later than eight days after the public installation of the new university rector (Rektorwahl), which marked the beginning of the new semester. The Rektorwahl was always on a Sunday, and classes would begin eight days later (a week from the following Monday).
Beginning with WS 1770/71, the private lecturers were also allowed to list their courses in the printed Verzeichnis; before that, they could announce them on the schwarzes Brett the announcement board on the outside of the university building, just to the left of the main door. It must be emphasized that these are simply lists of announced courses, and a fair number of the courses did not take place (generally for lack of enrollment).
Verzeichnis der Studiosorum were lists prepared at the beginning of each semester by the dean of all the students enrolled in that faculty. These were prepared only in the three higher faculties, as students weren’t enrolled as such in the philosophy faculty, and indicate the following (in Latin): the student’s name, age, residence, how long they’ve been at the university, lectures they attended the previous semester, and lectures they plan to attend the current semester. For example: F. A. Schultz, acting as dean of the theology faculty during WS 1755/56, had entered a list of twenty-one students intending to study with “Mag. Cant” (Magister Kant). The list was alphabetized by student name, and occasionally indicated the actual course; records were stored with the Etats-Ministerium [Arnoldt 1908-9, v.179].]
Mid-Semester Reports were lists entered some time near the middle or end of the semester of those courses actually taking place. Sometimes these come with a list of lectures to be given the following semester. There are also other drafts or versions of these reports. For instance, we find a record dated 29 August 1783 in Kant’s hand, listing all the courses for SS 1783 that he’s teaching (viz., logic, repetitorium, theoretical physics, physical geography), as well as a list of courses he intends to teach the following semester (metaphysics, pedagogy, repetitorium, natural theology, anthropology). It appears that professors would some semesters turn in their own lists at the faculty meeting, and other semesters enter the information into a common list.
Arnoldt reports that WS 1766/67 was the first semester for which these lists are fully reliable [1908-9, v.209]. As for when these reports are submitted, we have the following examples: SS 1766, entered September 16 [Arnoldt 1908-9, v.214]; WS 1768/69, entered March 18 [v.219 from the records of the Budget Ministry]; WS 1771/72, entered March 31 [v.235]; SS 1772, entered August 3 [v.236 from the minutes of the academic senate]; WS 1772/73 entered February 26 (and reported March 17); WS 1773/74, entered March 2; SS 1774, entered June 18; WS 1774/75, entered March 22; 1775/76, entered February 19; SS 1776, entered August 26; SS 1779, entered August 23; SS 1787, entered September 1.
The records for WS 1772/73 are hand-written and hand-drawn tables of five columns spread across two pages: (1) a narrow column indicating the discipline (e.g., “d. Philosophica”), written once at the head of the list; (2) “Soll gelesen werden”: a list of all announced courses, grouped by topic with headers inserted (e.g., “4) Metaphysics ... 5) Physica theoretica” with each course listed or a line with the professor’s names; (3) “Ist gelesen werden” (gives the time and whether it will be finished, or else dashes if it did not take place; (4) “Ist nicht gelesen werden”: gives reasons why the course didn’t take place (the standard reason being low enrollment); (5) an untitled last column for any additional comments. For example, it is noted here that Kant taught Anthropology from 9-10 instead of an announced physics course that failed.
End-Reports were prepared regularly beginning SS 1777 (?), to be included in the faculty senate records of all the classes that actually took place during the previous semester, with the following format: (1) The Faculty in which the course was offered (namely, Philosophy, Theology, Law, or Medicine), (2) the name of the teacher, (3) the name of the course, (4) the title of the textbook used, (5) whether the lectures were publice, privatim, or privatissime, (6) the number of students, (7) the date of the first lecture, (8) ... and last lecture, (9) if the course failed to meet, the reason for this (“causae quare non”). Copies of the end reports would be sent to Berlin as well as the budget ministry in Königsberg, along with a cover letter signed by the senators.
Other available sources of information are diaries and letters of contemporaries (for instance, Puttlich’s diary entries while a student and Hamann’s various letters mentioning Kant’s lecturing activities) as well as subscription lists [glossary] for the private lectures.
The least reliable of Arnoldt’s data may well be the beginning- and end-dates of Kant’s lectures (for each semester), since he seems not to have consulted calendars for the days of the week, and thus was less likely to notice either errant entries in the records or his own copy errors. Paying attention to the calendar, however, can provide helpful information. For instance, Kant routinely taught anthropology on Wednesday and Saturday (two hours back to back each of these days) during the winter semester, and physical geography the same time in summers. Yet he began teaching both of these courses on the four main lecture days (Monday, Tuesday, Thursday, Friday). So when did he change? Arnoldt appears oblivious to the days that Kant lectured on physical geography for SS 1775, even though he gives the beginning- and end-dates for this class, and these dates fall on a Wednesday and a Saturday, respectively [1908-9, iv.423]. Similarly, he is uncertain as to the days Anthropology was taught for WS 1775/76; but again, the dates recorded are for a Wednesday-Saturday schedule; to claim that Kant consequently did in fact lecture on these days requires only a modest inference.
 To whatever extent I have properly understood these materials is thanks primarily to the patient help of Werner Euler at Marburg, who enjoys a close familiarity with these records.
 Records associated with Kant’s physical geography lectures show some variability in the timing. The course announcement was submitted to the dean for WS 1755/56 on October 11 (classes began Oct. 13); submitted for SS 1757 on April 13 (classes began April 25); submitted for SS 1758 on April 11 (classes began the previous day), submitted for SS 1759 on April 28 (classes began April 30), submitted for WS 1759/60 on October 5 (classes began Oct. 8), submitted for SS 1761 on April 6 (classes began that day), submitted for WS 1761/62 on October 11 (classes began the next day), submitted for WS 1763/64 on October 10 (classes began that day), submitted for WS 1765/66 on October 13 (classes began the next day).
 Oberhausen and Pozzo  reproduce these printed catalogs from SS 1720 to WS 1803/4 and provide indices of (1) mentioned books of the Bible, (2) teaching faculty (prior to WS 1770/71, they included only the full and associate professors; afterwards, lecturers were included as well), and (3) textbooks, authors, and other persons mentioned in the course announcements. Since May 1736, a German version of the Vorlesungverzeichnis was published in the Wochentliche Königsbergische Frag- und Anzeigungs-Nachrichten (1717-74) [cf. Klemme 1994, 16], and beginning with WS 1765/66 also in the Königsbergsche Gelehrte und Politische Zeitungen (1764-96), e.g., WS 1765/66 (publ. Oct. 14), SS 1766 (publ. April 18), WS 1766/67 (publ. Oct. 17), WS 1767/68 (publ. Oct. 23), SS 1768 (publ. April 22), WS 1768/69 (publ. Oct. 10). An edict from Berlin of 2 March 1790, signed by Möllner [sic; typo in Arnoldt? Is Wöllner intended?], and addressed to the East Prussian Etats-Ministerium requested that public and private lectures be printed in separate lists in the Lecture Catalog “so that every poor student can quickly review them” [Arnoldt 1908-9, v.314]. The Catalog of WS 1790/91 was the first to list them separately, as can be seen in Oberhausen/Pozzo [1999, 575].
 Goldbeck [1782, 59] mentions this as a requirement of the theology faculty, but it would seem to be a requirement of all three. A list that I have examined of the medical students prepared near the end of WS 1785/86 includes the following: (1) name, age, birth place; (2) when matriculated; (3) the lectures now attending; (4) examination; (5) what lectures they plan to attend next semester; (6) any stipendia being received; (7) specimina (as respondents in disputations); (8) “ob mathesi und Sprache excolire” (whether they have mastered their mathematics and languages?).
 These are the reports often noted in Arnoldt [1908-9] as stemming from a Consess der Facultaet (i.e., a faculty meeting).
 There appear to be at least three genre of end-reports to be found in the records for the philosophy faculty.
Lectures announced and probably held are indicated with parentheses. In general, this status will be shared by all the courses given in any one semester, since either the list recording those courses actually given will be available for all the courses in a semester, or else for none at all. Sometimes there is other positive evidence, such as clearly documented student notes from that semester or other contemporary reports.
The main lecture days were Monday, Tuesday, Thursday, and Friday. Kant also routinely held certain classes on Wednesday and Saturday, such as physical geography and anthropology, and in the table I will always give the days for these two classes, when it is known. All the other courses met on the four main days, unless otherwise noted. The time of the lecture is listed when known (either in parentheses, if there is not positive proof that the lectures took place; otherwise without). The classes normally met for one “academic hour” (45 minutes) each day. Thus, most classes met four hours a week, and this is the case for all lectures for which a time is given but no meeting days. Anthropology and physical geography usually met on Wednesday and Saturday for two hours at a time, and the pedagogy lectures apparently met only one hour per week (on either Wednesday or Saturday).
The academic year was divided into two semesters by the Feast of St. Michael (September 29) and Easter, with the summer semester occurring from Easter to St. Michael’s and the winter semester occurring from St. Michael’s to Easter. St. Michael’s was a fixed date, but Easter fluctuates from year to year, since it falls on the first Sunday after the first full moon after the spring equinox (March 21), thus the date for Easter can be anywhere between March 22 and April 25, creating a possible difference in length between any two semesters of more than a month.
Each new semester began with the so-called election of a new Rector (Rektorwahl) for the university, taking place “on the Sunday after Easter for the summer semester, and on the next Sunday after the Feast of St. Michael’s for the winter semester.” This election — in actuality not an election at all, but rather the public installation of the full professor next in line for the position [more] — marked the official start of the semester, but classes would not begin for another week, since the distribution of the lecture catalog (Vorlesungsverzeichnis) was to occur “eight days after the election of the rector at the beginning of each semester” [Arnoldt 1746, i.199]. I provide the dates for the actual beginning of classes each semester (always a Monday) in the second column of the table.
Summer semesters are indicated by a single year (e.g., 1756), winter semesters are indicated by two years (e.g., 1756/57). These semesters were generally separated by vacations [more] of a month or more, as well as various holidays within each (vacation schedules changed according to the law of the state, but also were a matter of informal convention). When the beginning- and end-dates of classes are available, I list them below the name of the course, followed by the day of the week in parentheses (M = Monday, etc.). The beginning- and end-dates of repetitoria [glossary] always fell on Saturday, and so these days are not entered (in order to reduce visual clutter). In square brackets after these dates is the total number of lecture hours available, although from this number one must subtract any vacation time. These vacation dates have not yet been well-determined (see the discussion in Professors: The Academic Schedule), but in general we would need to reckon on one or two weeks at Pentecost and four weeks in August during the summer semester, and a Christmas vacation of four weeks during the winter semester.
I provide separate columns for Kant’s main lectures, and enter the rest in a single column. The exact titles given the various classes are found at Kant’s Lectures by Discipline.
The number of students attending the classes are provided behind a back-slash (“/” after the day and time). SS 1775 is the first semester for which these figures are available, when the beginning- and end-dates for the classes are also recorded. It is unclear how these student figures were determined possibly from subscription lists. In any event, they are perhaps best understood as rough indicators of actual attendance; any college teacher knows how the classroom population fluctuates, and typically drops during the semester. Several of the numbers are listed as approximations.
Finally, Kant apparently offered five additional but unnamed privatissima: 1767, 1769, 1769/70 (two), and 1770/71.
 From the statutes of the university at Königsberg, reported in Daniel Arnoldt [1746, ii.87].
 Werner Euler has brought to my attention a letter dated 22 September 1785 from the then-current rector, professor of law Georg Friedrich Holtzhauer (1746-1801) and addressed to the senate. Among other things, Holtzhauer noted that the lecture catalog needs to be distributed fourteen days before classes begin (per regulations of the Budget Ministry in Königsberg), and thus it needs to be posted next Sunday. The next Sunday would have been September 25th, and classes would have begun in 15 days on Monday, October 10. This regulation was mentioned merely in passing, as a reminder rather than as news.
Kant’s course on physical geography for SS 1757 was entered into the Faculty Minutes on April 13 (the Wednesday following Easter); for SS 1758, it was entered on April 1 (the Saturday following Easter); for SS 1759, April 28 (the Saturday after the election of the Rector); for SS 1761, April 6 (the 2nd Monday after the election of the Rector, and what should be the first day of classes); for WS 1761/62, October 11 (the Sunday after the election of the Rector); for WS 1763/64, October 10 (the 2nd Monday after the election of the Rector); for WS 1765/66, October 13 (the Sunday after the election of the rector).
[These concern lectures following SS 1775, after which enrollments and beginning- and end-dates were routinely available (we have figures for 34 of the last 44 semesters).]
Punctuality: Beginning with SS 1777, Kant always, or nearly always, began his four main courses on time, that is, Logic and Metaphysics on Monday of the first week, and Physical Geography and Anthropology on Wednesday of the first week. The first four semesters (SS 1775-WS 1756/57) he began Logic and Metaphysics three days late, and Physical Geography and Anthropology he began up to a week late.
The Practicals [glossary]: These always concerned the subject matter of the public lecture offered that semester (i.e., either Logic or Metaphysics), typically began the Saturday following the first day of classes during the 1770s, but after that usually began the second Saturday (of the 31 dates available, 19 are for one week later). Also, it usually ended the day after the end of the public lectures (e.g., Logic would end on a Friday, and the last practical would take place the following day, on Saturday), although there are exceptions: 1775 (one week later), 1780/81 (one week earlier), 1782/83 (17 days earlier), 1785/86, 1788/89, and 1789/90 (one week earlier). Attendance at the practical was, on average, one-fifth of the public lecture that semester.
Relative lengths of the main courses: Metaphysics would usually end after Anthropology, but Physical Geography would usually end after Logic (in nine semesters by one or two weeks). These four courses followed the semester end-dates rather closely, with the exception of Logic, which often would end early.
Physics, Moral Philosophy, Natural Law, Natural Theology, Encyclopedia: to the extent that we have data, they all began on Thursday of the first week of the semester.
Pedagogy appears to have always begun in the second week of the semester (and it was a one, or at most two, hour course).
Easter break was about eight days longer than the Michaelmas break, averaging 17 days for Michaelmas and 25 days for Easter. These breaks grew over the years, as well. In the 1770s, the Easter break averaged 23 days; by the 1790s it was 30 days. Similarly, the Michaelmas break grew from an average of 16 days to 23 days.
Semester Breaks: The average number of days between the last day of a semester of Logic lectures (at Michaelmas) and the beginning of the Metaphysics lectures was 23 days (vs. 27 days at Easter between the end of Metaphysics and the beginning of Logic). The average number of days between Physical Geography and Anthropology (at Michaelmas) was 19, and between Anthropology and Physical Geography (at Easter) was 32. Because the Easter break is about eight days longer than the break at Michaelmas, one would expect about that much difference between these averages. That there is instead a 13 day difference between the breaks between Physical Geography and Anthropology suggests that he ran his Physical Geography courses longer into the vacation. (For more on vacations, see The Academic Schedule.)
Available Lecture Days: When we compare the maximum number of possible class days for a set of lectures — namely, between the listed beginning- and end-dates (thus ignoring any midterm vacations) — we see that Metaphysics (in Winter Semester) averages 11 days more than Logic (in Summer Semester), yet Anthropology averages only 5 days more than Physical Geography. Similarly, when we compare courses meeting the same term: Logic averages 86 days and Physical Geography 89 days (in the summers); Metaphysics averages 97 days and Anthropology 94 days (in the winter). Variations in mid-term vacations may well equalize the overall length of the two semesters (we know only approximations of when and for how long these vacations lasted); but the differences between courses in the same semester should be constant regardless of these vacations.
 In this and the following averages, I compute the mean average for the semesters from SS 1775 through WS 1792/93. The following three years either are without data or else the data are skewed by Kant’s early ending of the semesters.