hksc logo
SearchPublicationsWebb CupWilde Cup

Jump to the top of the Article

Appendix - Directional Markings on Hong Kong Stamps & Covers [Summary]

See the 12 Items that have been contributed to the census. Can you add an item? [Updated:23 August 2007]

The aim of the Hong Kong Study Circle is to record and circulate information on the philatelic and postal history of Hong Kong and the Treaty Ports. Our main tool remains the journal which we publish quarterly and distribute to our members. The Internet now affords us a means of circulating information to a wider audience. We will occasionally publish, on our website, papers that we think are of interest to a broad range of postal historians and others. The aim being to encourage contributions of information from beyond our membership and to make more people aware of the interests of our members.

Are there yet more types of direction markings to be reported? I encourage readers to participate in the census. Send scans and measurements (in mm) to one of the email addresses given below. Ask a friend to help if you do not have a scanner.

The following e-mail addresses are displayed as graphics so that Internet marketers cannot automatically copy them to their bulk mailing lists.
Bulletin Editor, John Wilson
Webmaster, Michel Houde

Direction Markings:
Re-Direction Markings Reconsidered

Charles A. Jones

[First published in the Journal of the Hong Kong Study Circle, April 2007 No. 341]


Names of ports, such as Hong Kong, Bangkok, and Singapore, are occasionally found as unframed, straight-line markings on loose stamps, post cards, and covers of Hong Kong, Straits Settlements, and one example so far from North Borneo. These markings are on issues of the first three reigns, but have not been reported on KG VI issues. Accompanying figures show examples of these markings.

These markings are not common, but they exist in surprising variety. A recent census revealed 22 different types on Hong Kong stamps, cards, and covers; nine on Straits Settlements, and one on North Borneo. Results of the census are tabulated in the appendices to this article.

This article addresses the types of marking, and whether they all served the same purpose.

Background. I know of two previous references to these markings. The earliest is in HKSC Bulletin 35 (August 1956). Mr. Douglass of San Francisco wrote that he had four examples of these markings and asked about their purpose. The answer (which Col. Webb may have provided) was that ?these are redirection marks, usually to be found on covers and cards addressed to naval or merchant ships which have moved on? [emphasis added].

Evidence for this interpretation was a card mailed in 1895 from Colombo to a stewardess [sic] onboard a German boat at Aden. The writer stated that the word SHANGHAI was stamped on the card, and that the card was twice forwarded, first to Hong Kong and then to Shanghai. No illustration of the card or other details were provided.

The second reference is Webb (p. 152), where he gave these markings only brief mention under the heading Early Ship Re-Direction Markings. He reported 1) Shanghai stamped on a cover, 2) Hong Kong stamped across the face of stamps from ?Ceylon, Straits Settlements, etc.?, and 3) Singapore, Penang, Colombo, and Rangoon on Hong Kong stamps. Webb did not illustrate any of these markings. He interpreted them as,

?[markings] applied by the post office of the place concerned to correspondence to the crews of ships which had already sailed when the letter arrived, and were an indication that the letters were re-directed by the post office to another port of call.

Neither Mr. Douglass? card nor the Shanghai cover mentioned by Webb were reported during the recent census.

In the case of both Mr. Douglass? card and the Shanghai cover mentioned by Webb, the conclusion was that the purpose of these markings was to send correspondence chasing after a ship that had sailed to another place. In other words, these markings (by whatever name they are to be called) were forwarding markings. With the possible exception of an 1881 cover discussed below, none of the markings in the current census support the conclusion that they are forwarding markings. It is possible of course that these markings may have served more than one purpose, from time to time, as needs arose.


To date, the census comprises 37 items. Of these, 27 are Hong Kong (and Treaty Port) usages, nine are from Singapore, and one from North Borneo. Among the 27 Hong Kong items, five are duplicates for a total of 22 different types. There are no duplicates among the nine Straits Settlement usages.

I conducted the census among members of the Hong Kong Security Markings and Perfins (HKSM&P) Study Group, six of whom participated in the census. Thus, on one hand, the sample size (number of collections represented) is small and almost certainly does not represent the whole population of these markings in other collections worldwide. On the other hand, the participants in the census are all long-time collectors with extensive and respectable collections of cancellations and markings. In any case, conclusions I draw from the census are subject to change as more of these markings, especially those on cover, are reported.

In all cases, participants in the census provided scans or photocopies of the markings in their collections. Many of these can be seen at the HKSM&P website, which may be found at: Although these markings are on the website, this is not to say they are security markings, although some may be.

All markings in the census are listed in the appendices by name (port, destination), then by type based on measurements and characteristics of the letters. Markings on Hong Kong stamps, cards, and covers are in the first table; markings on Straits Settlements and on one stamp from North Borneo are in the second and third tables. My discussion of these markings focuses on those from Hong Kong and the Treaty Ports, but nothing about the markings on Straits Settlements or North Borneo suggest they are different.


The markings fall into two populations: those on loose stamps and those on cards and covers. None of the types on stamps have been reported so far on cards and covers, nor have any of the types on cards and covers been reported on loose stamps. I can not explain this dichotomy. It may or may not be an artifact of sampling (small sample size), but it does raise the question whether all of these markings served the same purpose.

The markings on loose stamps are all more or less centrally struck, as if clerks were instructed to place them there. (This assumes the markings are official and not private markings.) On cards and covers, however, the markings are, with two exceptions, struck well away from the stamps in almost any convenient white space. The two exceptions are an 1881 cover and an F1 letter, both very early examples that predate later examples by about a decade. [The F1 letter is undated but most likely from the early 1880?s. The F1 obliterator was probably withdrawn in late 1884 or early 1885; the latest F1 cover recorded by Gurevitch is December 1884. The next oldest (datable) markings in the census are an 1893 cover (Hanoi Type 1) and a loose stamp of 1894 (Hong Kong Type 1.)]

Loose Stamps. Of the 22 types on Hong Kong stamps, cards and covers, 11 types are on loose stamps. One of the nicest (Journal 340 cover and Figure 1) is a complete strike of BANG KOK (as two words with space between G and K) on a re-joined pair of the 5 Crown CA issue (Hong Kong Type 1).

Figure 1. BANG KOK, Type 1, as two words, c. 1890?s,
on 5 blue, rejoined pair.

The RPS certificate that accompanies this pair states that the marking is a ?? semi-official postal marking used on mail with non-English writing to assist with sorting?. In other words, in the opinion of RPS, the marking was a post office marking, official or semi-official, used to help non-Chinese reading clerks sort mail addressed only in Chinese. This explanation is significantly different from Webb?s conclusion that these markings were forwarding (as opposed to sorting) markings.

The expression semi-official on the aforementioned certificate is somewhat ambiguous. I take it to mean that the marking was applied by British post office functionaries in response to an official instruction or practice for which no written instruction has been discovered.

Covers. Regrettably, markings on loose stamps don?t reveal their purpose. Markings on cards and covers are more useful in this respect.

Of the 11 types on cards and covers, nine are on red band covers addressed only in Chinese, one is on a 1918 card addressed in Chinese, and one type is on both the 1881 cover addressed in English and the early 1880?s(?) F1 letter addressed in Hindi.

Figure 2. Dual language FORMOSA, Type 1, on post card, 1918.
Note Chinese characters below FORMOSA.

The 1918 card (Formosa Type 1) (Figure 2) is the latest datable marking in the census. This marking is a dual-language type with both English letters and Chinese characters. It is the only dual-language type with Chinese characters below the English destination.

The 1881 cover and F1 letter are early examples, as explained above, and may represent special cases, as explained below.

Figure 3. SAIGON, Type 1, on 1903 red band cover.
Saigon is lined out for no apparent reason.

[?Binh Tay? in Roman letters is the name of a major Chinese market within Saigon. ?B.Ed.]

Figure 4. HANOI, Type 1, on 1893 red band cover

The Saigon Type 1 (Figure 3) and Hanoi Type 1 (Figure 4) covers are typical of the red band covers, all of which are addressed in Chinese. The address, written in Chinese characters, is their significant commonality. In no case is there evidence, such as a lined-out first address, to indicate the card or cover was forwarded to a new address.

Dual Language Markings. Several of the markings reported (seven of 22) include the name of the port (or destination) in both English and Chinese. These markings are in two lines. With one exception (Formosa Type 1), the Chinese characters are above the English name. Chinese characters read from right to left following Chinese custom. The Rangoon Type 1 marking (in Figure 5) is an example. The character ?GON? appears above the ?A? of Rangoon. If the full strike can be found, the character for ?RAN? should appear above one of the O?s of RANGOON.

Figure 5. RANGOON, Type 1, 1898; and two dual-language examples:
SINGAPORE, Type 2, 1901, and BANGKOK, Type 3, c. 1900?s.

Figure 6. SINGAPORE, Type 5, 1901.Figure 6. SINGAPORE, Type 5, 1901.

The red band cover to Singapore (Singapore Type 5) (Figure 6) is another example. (The dark line along the left side is the red band that overlaps from the front of the cover where the address is written in Chinese.) The illustrated side (back) of the cover shows a 1901 Hong Kong circular datestamp. The cover entered the post at Hong Kong, where I suspect the Singapore direction marking was struck. Two persons, both fluent in Chinese, one of whom is a member of our study circle, inform me that the Chinese characters do not read Singapore. They suggest the characters may represent an old name for Singapore. (Further comments on this would be appreciated.)

The Hongkong Markings. There are two types of Hong Kong markings in the census. The first (Hong Kong Type 1) (Figure 7) is the word Hongkong (one word) in capital and lowercase letters. This marking may have been applied at the Hong Kong G.P.O. during the sorting process to indicate local delivery of mail addressed in Chinese. (This interpretation assumes the marking is not a private or security marking.)

Figure 7. HONGKONG, Type 1, on CCA issue, c. 1901 or later.

The second type (Hong Kong Type 2) is on both the 1881 cover and F1 letter previously mentioned. On each of these, Hong Kong (apparently as two, closely spaced words) is struck across the face of the stamp. These are the only examples in the census of a card or cover with the marking across the face of the stamp. Both of these letters predate the later markings by about a decade, and therefore may have served a different purpose from the later markings.

Figure 8. HONGKONG, Type 2, on missent(?) 1881 cover addressed in English to Bombay.

The 1881 cover (Figure 8) is exceptional in other respects. The letter is written and addressed in English script. It is dated internally, 2 June 1881. The letter bears a Bombay receiver of OC 17, without year date, but presumably also 1881. If so, the letter was in transit four months, long enough for considerable traveling.

The writeup that accompanies this letter gives the following interpretation: The letter was missent, perhaps to Australia, found there in the mail and redirected back to Hong Kong and thence to Bombay. If so, the Hong Kong marking would have been applied at Australia (or wherever) to direct the letter back to Hong Kong as perhaps the most direct or timely route to Bombay. Regrettably, there are no notations, receivers, or other markings to docket the letter?s wanderings, first to a wrong destination and then its transit back through Hong Kong to Bombay - if this indeed happened. However, if the foregoing interpretation is correct, the Hong Kong marking would be a forwarding or re-direction marking along the lines of Mr. Douglass? card and the Shanghai cover mentioned by Webb.

Figure 9. HONGKONG, Type 2, on F1 folded letter
addressed to Hong Kong in Hindi. Early 1880?s?

The F1 letter (Figure 9) is undated, but probably dates from the late 1870?s or early 1880?s, based on usage of the F1 cancel. The letter, addressed in Hindi script, is said to be addressed to Hong Kong. It is a local letter. Faint pencil writing across the horizontal double lines appears to read Hong Kong in English scribble. The Hong Kong marking ties the stamp, as it does on the 1881 example.

It seems likely that the Hong Kong Type 2 marking was to facilitate sorting of mail, addressed as it is in Hindi, a form of writing likely to have been inscrutable to many English or Chinese speaking mail clerks. The letter is not forwarded.

Security Markings(?). I cannot rule out that some of these markings, especially those on loose stamps, are security markings. Security markings usually identify the firm or owner by names and initials. But some security markings are anonymous, e.g., lines drawn across the face of the stamp or the word Stamped, in various forms, applied to the stamp to prevent theft and resale (HKSM&P website, Group IV, Miscellaneous Markings).

A case in point: two examples of the name Amoy, struck in violet, recently came to light (Figure 10, and HKSM&P website, Group 1, #105d). More recently, a central strike of this Amoy marking was discovered on a loose stamp. It is now known that the violet Amoy marking is part of a security chop belonging to Pasedag & Co, a firm in Amoy.
This AMOY security marking, when it appears alone and not coupled with the full PASEDAG & Co/AMOY security marking (in two lines), may look like a direction marking. Therefore, I must admit the possibility that some of the markings on loose stamps in the census may be security markings.

Very few security markings are dual language with both English and Chinese writing. Only one of these bears resemblance to the dual-language markings identified during the census and discussed here. It is the entry on the HKSM&P website under the heading, What Is It?. (On the website, look for the word TECK with Chinese character above.) This said, none of the markings in the census seem to me to be security markings.
The picture might change if a Hong Kong or Treaty Port cover can be found, addressed in English, not forwarded, and with one of these markings struck across the face of the stamp in the manor of a security marking. Such usage would almost certainly be as a security marking.

What Else Is Out There? A newly discovered cover (Figure 11) sent from Amoy to Batavia is interesting and different in one respect from other covers reported in the census: It is the only cover addressed to an expatriate Chinese firm in both Chinese characters and with a handstamp in English letters.

Figure 11. BATAVIA, Type 2, on 1894 cover Amoy-Batavia, showing Amoy cancel (overstruck HK or Singapore), transit marking of N-I Agent Singapore, 30 January, and final Batavia cancel.

The cover, a red band cover, has three handstamps on the back: FRANCO (?post paid?), BATAVIA, and ENG.KIM. All three are struck in slightly different shades of red. The letter is addressed to an individual, a Mr. Hung, at the Eng Kim firm, Batavia.

I believe the Batavia marking is a direction marking, as discussed here, struck either at Amoy or Hong Kong. I believe the ENG.KIM marking is a private marking, most likely applied by the writer (or the writer?s firm) at the point of origin, Amoy. The Eng Kim marking would not be a security marking because it is struck well away from the stamp.


The following observations are based on the census of markings that occur on Hong Kong (and Treaty Port) stamps, cards, and covers. Nothing about the markings on Straits Settlements and the one North Borneo example suggests they are different.


  • So far there are 22 different types of direction markings on loose stamps, cards and covers of Hong Kong and Treaty Ports.
  • Surprisingly (to me), the census tallied only five duplicates. All others were one of a kind. More duplicates, as well as new types, likely exist.
  • Most are struck in red or orange-red (19). Fewer are black (3). All but one marking from the QV era are red. All KE VII and KG V examples are black.
  • Twelve (12) of the markings are in English, 7 are in English and Chinese (bilingual), and 3 are incomplete and indeterminate in this respect.
  • Most of the English markings are in block letters with or without serifs. Three (3) are in script.


  • The markings were used during the first three reigns. Markings from the QV era are most common.
  • The earliest datable example is 1881; the latest, 1918. The 1881 cover, however, may be a forwarding (re-direction) marking. Excluding the 1881 cover and F1 letter, the earliest datable marking is 1893 (Hanoi, Type 1).


  • The census identified 10 destinations: Bangkok, Batavia, Canton, Formosa, Hanoi, Hong Kong, Penang, Rangoon, Saigon, and Singapore. (Webb mentioned a Colombo marking, but this marking was not reported during the census.)
  • All destinations are in the Far East, Southeast Asia, or the Malay Archipelago where there were expatriate Chinese communities.

Two Populations: Loose Stamps v. Cards and Covers

  • Markings are equally represented by examples on loose stamps and on cards and covers (11:11). This ratio will likely change as more collectors participate in the census.
  • There seems to be two populations of these markings. None of the 11 types on loose stamps have been seen on cards or covers. None of the 11 types on cards and covers have been seen on loose stamps. This opens the question whether the markings on loose stamps are security markings rather than direction markings.
  • The dichotomy between these two populations continues in another respect: Types struck on loose stamps tend to be more or less squarely (centrally) struck across the face of the stamp. This is so consistent that it seems purposeful and not accidental or happenstance. Conversely, types struck on cards and covers are, with the exception of the 1881 cover and F1 letter, all struck well away from the stamps (or indicia).

Addressed in Chinese (or Hindi)

  • Among the 11 cards and covers, all but the 1881 cover and F1 letter, are addressed solely in Chinese (and one in Hindi). True, the Saigon cover bears a handwritten English translation of the addressee?s name and location; but I suspect this translation was provided by a second party, not necessarily a postal clerk.
  • None of the cards and covers in the census have additional markings or second addresses to indicate they were forwarded (re-directed). This includes the 1881 cover that seems to have been forwarded.

Security Markings?

  • Yet to be explained is the fact that the markings on loose stamps are different from those on cards and covers, as explained above under the heading, Two Populations. This is puzzling, but may only reflect the small sample size of the census.
  • The markings on loose stamps may be security markings but there is no direct evidence from the census to support this. Of the over 1000 different security markings known to exist, only one that I am aware of bears any similarity with the markings in the census, and none of the markings in the census have been considered security markings by the many who have collected security markings for so many years.


  • From the census, I conclude that direction markings are a type of instructional marking to facilitate sorting of mail addressed in a language other than English (e.g., Chinese or Hindi). It was not the purpose of these markings to forward mail to a second address after attempted delivery to the first address.
  • Forwarding (re-direction) markings, as Webb described them, may exist (e.g., Mr. Douglass? card, the cover Webb described, and the 1881 cover discussed above); but the markings in the current census appear to have served a different purpose. They deserve their own designation: direction markings (without prefix).
  • Re-direction is an inappropriate term for these markings because the prefix re- means to turn back or repeat, neither of which direction markings were intended to do.


  1. Are there yet more types of direction markings to be reported? I encourage readers to participate in the census. Send scans and measurements (in mm) to one of the email addresses given at the top of this page, click here. Ask a friend to help if you do not have a scanner.
  2. Where are Mr. Douglass? card and the cover Webb described? It would be useful to confirm that these two items are indeed forwarding (re-direction) markings and not direction markings, as defined here.


I thank members of the Hong Kong Security Markings and Perfins Study Group who participated in the census of these markings. There would be no census without their time and interest in this endeavor. I also thank Ming Tsang, Hong Kong Stamp Society, and Dr. David Ko, China Stamp Society, and Timothy Summer, for help with Chinese translations. I am in debt to Lee Scamp for his encouragement of this project and his review of various drafts of this article. Opinions and factual errors are of course my responsibility.

Are there yet more types of direction markings to be reported? I encourage readers to participate in the census. Send scans and measurements (in mm) to one of the email addresses given at the top of this page, click here. Ask a friend to help if you do not have a scanner.

Are there yet more types of direction markings to be reported? I encourage readers to participate in the census. Send scans and measurements (in mm) to one of the email addresses given at the top of this page, click here. Ask a friend to help if you do not have a scanner.


Appendix - Directional Markings on Hong Kong Stamps & Covers [Summary]

See the Items that have been contributed to the census. Can you add an item?

Home | Membership and Fees | HKSC Bulletin | Meetings Scheduled | Links Publications | Webb Cup | Wilde Cup

Hong Kong Study Circle