Masson’s Dating

David Parkes Masson (1847-1915) was a most assiduous and careful collector of Kashmir covers “as they were happening.” His dating conversions from the Persian are often seen in his distinctive hand in red ink. Some of his conversions are defiantly at odds with the evidence of the British date stamps. It leaves us with a little puzzle that is hardly short of disturbing.

There are many cases in which Masson’s conversion puts the despatch date after the delivery date. Scan left: Our utility converts “11th Assu” to 25 September, with comfort from the cds. The inconsistecy by almost a week was certainly evident to him, but it clearly did not concern him for some good reason. What was it? Detail, Lot 183 Sturton Sale.

Examples of Masson’s datings are gathered below (data collection for now). One feature that shows up immediately is that the discrepancies are of markedly different duration, from near perfect agreement to differences one way or the other of (so far) up to 16 days. We begin to feel that these shifts are an artifact of a parallel lunisolar convention, which Masson wished to keep tabs upon for some reason, whereas the postal system adhered to the pure solar convention. How we describe the discrepancies in our table can be seen in the following examples:

For the detail on the left, our conversion utility gives 18 November in place of Masson’s 27 November, and we call that a “+9 day” difference because Masson’s reckoning is later by 9 days. Similarly, the detail on the right converts to 21 April on the utility, thus a “−2 days” would be recorded in the table.

On these two examples, Masson’s conversion dates were 6 and 10 days past the delivery date as given by British date stamps. There are several such in the Masson record. Another little peculiarity shows up in the table below with the pair of covers having Masson assignments 16 & 17 Phagan 1924. He does convert these to successive days, but backward running: 26 & 25 February 1868, respectively.

Masson’s “Chet” and “Jeth” are our ćait and jeţh, respectively. Dates in manuscript become scarce in the New Colors period, and we have yet to see a Masson at all. So far, the signed average Δ is about +4.1 days and the unsigned average is about +6.8 days (not including the shift of year marked in boldface in the table, which is caused by our use of baisākh as the advent month of the year, not ćait). The collection of dates in the table is yet too sparse for obtaining an instructive graph; we should one day like to see how the two systems go into and out of phase with each other. Perhaps there are not enough covers extant ever to do so.

By the way, in cursive writing ćait and jeţh can be difficult or impossible to tell apart. Both are often rendered by a simple undotted hook and completed with an uninterrupted flourish. On the left is ćait 1936, and on the right is 2 māh jeţh 1933, both with cover corroboration. Other pairs of months can be confused as well in rapid cursives, such as even māgh and baisākh.

Appendix: Some Lunisolar Complications

In the hybrid lunisolar calender mayhem occurs when two new moons occur within a solar-designated month, something that occurs about once every three years. It seems in effect to cause the afflicted month to “start early,” sometimes considerably so. We quote the alarming lines from Platts p 398 quoting Forbes: “Hence although the month baisākh begins de jure about the 11th of April, it may have commenced de facto from one day to twenty-eight days sooner,” depending on when its first full moon occurs. Other complications arise when accurate daily conversions are to be achieved for a particular location. If you lurk in dictionaries you might also have encountered terms referring to certain intercalary months that are potential disturbances to the system, such as bīsondh and adhimās ~ half-month. The former is of 20 days duration in which money-lenders did not charge interest; the latter, usually of 11 days, comprised the omitted days between the end of the lunar year and the start of the solar year.

Another complicating issue comes in another footnote (Masson II, page 1):

“. . . In some cases I may still be fourteen days ‘out’ in transposing the Christian dates for the Hindu ones. This is due to the Hindu month being divided into halves, shudi and badi, and the envelopes not showing in which half of the month the letters were written. Thus an envelope may bear the date 1st poh 1923; this might mean 1st poh badi, corresponding to our 22nd December 1866, or 1st poh shudi, answering to our 6th January 1867.” By the way, our utility conversion for 1 poh 1923 is 14 December 1866, which represents an eight-day discrepancy in itself.

For the item shown above, the Masson date discrepancy changes (one of) the accepted advent dates for an important stamp, the ½a black Kashmir single die. The conversion utility assigns ‘9th Assu 1923’ to 23 Sep 1866, where Masson’s 3rd October is ten days later. That dating is apparently on the reverse, likely in javab notation, and represents the letter’s pickup date. The stamp might therefore have been affixed even in mid-September or earlier. Haverbeck Sale Lot 1374; scan Séfi & Mortimer Plate 11.

To top of page