Resources > Identification Help

Isegrim Really Works -- Update and Encore

(1/6) > >>

*** Thanks to Helvetica / Dane at Wildwinds, Isegrim is at least halfway back as of two days ago, and in many ways better than ever; latest news at, Reply #55 (Isegrim now reloaded in xls- form; many fixes and tricks in the present essay could now use overhauling as well). Note that older links no longer function. ***

Isegrim Really Works / Really Working with Isegrim: A Short User's Guide to a Great Ancients Database

I was invited to write up this guide as an article for the Articles and Resources department on the main Forum website, but this board is as useful a venue as any; I know this set of ground-rules has always worked really well for me, and I hope other members will now find them useful as well. Isegrim is an excellent fully searchable union-archive of numismatic sources for ancient Asia Minor*, one of the richest and most diverse regions in ancient numismatics (even if your coin isn't from Asia Minor, this resource can help you confirm that, and thus save you time); Isegrim is unfortunately also a terrible tease, to the point that at times you can't find it online, and at times it appears you can't use it to find anything like the entries you want.  I propose to fix all that, with demos.

Logging On

The current anonymous index-login is here:

Notice the the name ISEGRIM is blue text and links directly to Isegrim whenever it is typed in the Classical Numismatic Discussion. The links are generated automatically.

You may need to refresh the page. Once you've arrived at a functional login page, simply click on "anonymous, English," and then choose option "Search Asia Minor," and so off you go. **


NOT ENOUGH OR TOO MUCH? Start with fewer terms and add more if you get too many results.

SPELLING MUST BE PERFECT. Geographical names usually use Latin spellings. UK English spelling is used vs. American spelling. 

ENTRIES MUST MATCH EXACTLY WHAT ISEGRIM HAS IN THE DATABASE, WHICH IS NOT ALWAYS WHAT YOU EXPECT. For example, the emperor Severus Alexander is listed in Isegrim as Sev. Alexander. Searching for Severus Alexander will not find any results.

USE THE WILDCARD (.*) BEFORE AND AFTER SEARCH TERMS. Note the wildcard is a period followed by an asterisk, not just a plain asterisk.

Isegrim has a powerful but non-user-friendly search-engine that often acts less like a guide than a bouncer, shutting you out completely and ruthlessly; it demands exact input in order to rule out irrelevant entries, but there's quite often clearly an arbitrary side to what form the exact input takes (more on this in Reply #1). You can thwart the exactness-requirement with an open or wildcard extender (= ".*") at either end of any Isegrim input, but too much of this means that you'll probably end up with far too many entries for comfort.  Fortunately once you obtain any relevant entries at all (Use just one or two key specifiers to start with, and always the simpler the better, once you get the knack), you can learn as you work, the way I did, and so after just one more quick note on the basics we can get down to cases and prove the point.

Isegrim Formats (Output and Input)

Every useful specifier or "descriptor" in an Isegrim entry is distinguished with a two- or three-letter prefix; you can use most of these as search-inputs as well, for example one or more of the following (although here, too, the simpler the better for starters):

vs:.*antwn.* vt:caracalla va:beard rt:zeus m:ae zit:.*3099.* po:saitta pro:lydia

[ vs = obverse legend, in this case a mere fragment supplemented with two wildcard extenders .* .* (note the great Partial-Legend Search-Resource this also presents); vt = obverse type; va = obverse attribute; rs = reverse legend; rt = reverse type; m = metal; zit = collection or auction reference; vgl = compare reference; vg or rg = obverse or reverse countermark (= German "Gegenstempel"), where present; po = issuing city; pro = issuing province, the latter two classes of terms typically Latinized in Isegrim; emperors' names are moreover quite often peculiarly shortened, for instance ant. pius, marc aurel, or sept. severus, so that you may do better with a wildcard-extender and a still shorter form, for instance .*pius ]. Other specifiers in Isegrim entries are equally worth knowing but need further tweaking to work as search-inputs (more on this in Reply #1), for example gew -- weight (German "Gewicht") and gr -- size (German "Grösse").  Some of Isegrim's transliterations of Greek legends are surprising: Theta or :Greek_Theta_2: -> T', Upsilon or :Greek_Upsilon_2: -> Y, Phi or :Greek_Phi: -> P', Chi or :Greek_Chi: -> C', Psi or :Greek_Psi: -> P'' (two apostrophes, not one quotation-mark). Once you've caught an array of results (open-wildcard extenders will make doing this a lot easier, but with too many wildcards your catch will max out at 500!), you can easily search the array on your screen just by keying [Control-F] in Firefox or Explorer. Isegrim as it currently stands is a text-only resource, but with its standard catalog-references searchable both in actual imprints and on other sites, you'll be able to access a huge range of plate-coins both online and off, a coin super-resource on your desktop.

Demo #1 (see first photo) -- Weeding Entries With Isegrim (Re: "Tiberius and Livia")

Discussion topic 38652

--- The inscriptions or legends on this coin certainly aren't that legible, though what's left on the flip-side (t'ea sebas ...) does help clinch the ID with the RPC entry (RPC I 4049, actually not Tiberius but Augustus) that I noted in that earlier posting, backed up with a scan.   Here's how I ended up there:

Isegrim search #1: vt:tiberius rt:livia -- The Mytilene coin this search turns up lacks a feature the question-coin has (inconspicuously radiate male portrait).  So I thought I would search a more basic distinction of our question-coin = a left-facing male portrait (obverse) and a right-facing female (reverse).  Search #1 showed how Isegrim wants these described***, and so on to Isegrim search #2:  vt:portrait man l rt:portrait woman r.  This turned up just nine types, and just one had a matching (radiate) male portrait along with a matching reverse legend.

Demo #2 (see second photo) ---

Discussion topic 41650

Searching only what's clear even if not especially distinctive on this uneven off-center coin

vs:.*ney.* va:beard.* rt:eagle rt:bird fr [= bird front, or bird facing]

we arrive at a few very similar specimens, all of them from Apameia in Phrygia.

Demo #3 (see fourth photo) -- Info request: Severus Alex from Tarsus

Discussion topic 43379

A very rare reverse; just two specimens cited in Isegrim (arrived at via search-string vt:.*alexander rt:nike po:tars.*); a quick visit to "Search Bibliography" (see below) and then a second quick visit to Google for "Cox 'Numismatic Notes and Monographs' 92" supplies us the full reference for the second specimen = "Dorothy H. Cox, 'A Tarsus Coin Collection in the Adana Museum,' Numismatic Notes and Monographs 92, 197" -- not too bad for a couple of minutes at the keyboard!

* Dane's new xls version includes something not far from a union-bibliography for Isegrim. The original site referenced virtually all useful imprints through 1985 and a few other substantive imprints through 2003 (more on this in Reply #1); a good sense of the range of the sources included may be gleaned from the plethora of entries forthcoming for any broad search, for example vt:herakles rt:lion.  The database ranges from the earliest of Asia Minor coins (excluding Cyprus) to the latest provincials, but excludes Alexander the Great, for whose coinage see esp. the Price-keyed Wildwinds entries at

Wildwinds Alexander the Great

and the more general online pictorial resources (esp.,, [partly offline, but most entries accessible via Wayback Machine], RPC for most emperors, the British SNG, the Recueil Général, Imhoof-Blumer online, and the Weber Collection) at

Discussion topic 32513

At this stage it bears noting that Isegrim's place-names may need to be de-Latinized if you want to access corresponding materials elsewhere online: Cius -> Kios, Cyme -> Kyme, Cos -> Kos, Coracesium -> Korakesion, and so on. 

** If you are indeed here to search Asia Minor coin-types (at this point you could also choose Isegrim's "Search Bibliography," which unpacks cryptic abbreviations), **The Easiest Way To Get Started With Isegrim Searches** is to cut-and-paste one of the search-strings (in bold) I include in this thread, starting vs, rt, or vt; once you get any relevant entries at all, you can fine-tune your search along lines I explore in the demos. 

*** If these were non-historical figures -- mythic persons or personified abstractions -- the inputs would read instead vt:head man r rt:head woman r, where as always in Isegrim searches the multiple inputs involve an invisible "and" (really = "both ... and"); more on classification-distinctions like "head" vs. "portrait" at the end of Reply #1.

Here's a bulletin-board which I'll update at need for my earlier posting on Isegrim:

--- There is after all a way to search by weight (gew) or size (gr) as well as by minting-dates (pz) in Isegrim: along with other relevant descriptors, for weights enter a lesser and greater value -- gew1:3.72 gew2:3.82, for example -- which are processed as lower and upper weight-limits and give you everything in between.  (If any specimen in a particular entry has a weight that's within your weight-range, it will show up as part of your search results.) Something similar holds for gr1 / gr2 inputs (see new demo) as well as for minting-dates (pz1 / pz2); from the updated English introduction at:

Isegrim - English Introduction

followed up with a trial-run or two, I've concluded that this is the gist of it: you can input one year, e.g. pz:18, in the hopes that it corresponds closely with the date of the entry you're seeking, or else you can enter an earliest and latest, e.g. pz1:10 pz2:60, and get everything that was probably minted in that range for the other descriptors in your search-string.  MAJOR CAUTION: Being copied from a great variety of sources, not all Isegrim entries include weight, size, and minting-date specs (though they generally do give at least one or two), so a targeted search with particular weight, size, or minting-date specs may miss coins that are actually in Isegrim. By a similar token, you may miss coins included in Isegrim if your input includes legend-inputs that do not allow for any variants or faults in the legends recorded in Isegrim. HELPFUL HINT: Subject to the potential exclusions that I just now mentioned, you can also employ bounded searches of these various sorts to cut down a broad yield of results to a manageable size; while a very broad search, for example vt:.*os, yields too many results to display, vt:.*os gr1:22 gr2:24 yields 470 results, just below the 500 truncation-point.

New demo for size-searches (scan posted below)
The coin pictured below is a badly misidentified Caracalla AE25 -- "Artemis and Apollo from Seleucia [!]" -- that I recently purchased on Xbax.  Searching Isegrim with the search-string vt:caracalla rt:heads 2 rt:head man l gr1:24 gr2:26 (vt -- obverse type; rt -- reverse type; gr1 -- boundary-size #1; gr2 -- boundary-size #2) we get Hygieia and Asklepios from Eirenopolis, Cilicia, which the legend confirms is correct.

--- Isegrim includes virtually all Asia Minor listings from RPC I and from Lindgren III ("SLG LINDGREN III"), so its coverage extends down through 1992 and 1993, when those volumes were printed, and somewhat beyond (just two listings from SNG München (Munich) 20 [1995] and only intermittent listings from RPC II, printed 1999; I have noted one listing corrected in 2001, a single auction referenced from 2002 [AUKT GORNY 118 1720], and one study repeatedly cited from 2003 [ZIEGLER AMS 2003]); many hundreds of entries from SNG PFALZ 6 = PfPS 6, also published in 2001; you can test other SNG coverage by searching zit:.*SNG X.* where X stands for the rest of an SNG title, for which now see the SNG titles below in Reply #9).   With the search-entry zit: or vgl: (for "Compare") and a catalogue-entry in form "RPC I 2417.*," for example, you can summon up virtually everything else that's included in a given printed entry from RPC I; Isegrim is thus also a digital RPC I for Asia Minor, and a digital version of whatever else Isegrim incorporates, though its uptake of RPC I seems especially thorough and painstaking.  Isegrim includes some coins not others for marginal figures and mints, for example, "Rhoemetalces" or "Cyprus"; here you're definitely better off using alternative sources, not mainly relying on Isegrim.

--- Field-insignia (VF, RF) and countermarks (VG, RG) can be very useful in worn-coin ID's, but the Isegrim entries for all these descriptors have quirks of their own; a good way to explore them is to run open searches with each of the pertinent descriptors (VF, RF, VG, or RG:.*), and then see what you get.  Letter-content appears with a prefix intended to sort letters, values, and dates ("LET AS," "VALUE I," "YEAR GXR" = 163); letter-content in VF or RF will also appear in VS or RS, but without the fresh prefix (thus RF: YEAR GXR but RS: P'LAYIOPOLEITWN ET GXR). Here is a more or less typical output for an obverse countermark arrangement; note that each of the terms constitutes a good input in its own right:


"Round," "oval," and "angular" are often but not always used to distinguish the shape of a countermark-strike; in this case repetition of "round" means we actually two obverse countermarks on the same coin, with "head man l" and "emperor" both referring to the second round countermark's subject or content.  These descriptions can vary even for the same countermark; thus these all seem to be the same countermark on coins from Lydia ad Sipylum, Magnesia:


Numbers in parentheses found in some of these entries indicate which specimens out of several described display which field-insignia or which countermarks; these distinctions are not always clear and not always complete, but still often help fix a coin's provenance.  Isegrim reports some but not all of the Howgego numbers for cited Greek Imperial countermarks; Howgego can thus complement Isegrim and vice versa, since Howgego does not report pre-Imperial coins which quite often share countermarks with later issues.

--- If you are searching particular divinities' names it's important to follow the name with a wildcard-extender, for example "zeus.*; without that, the search won't include special cult-aspects of Zeus, for example Zeus Kelaeneus or Zeus Lydios (other random examples: Artemis Ephesia or Hekate Triformis, and remarkably also Zeus Sarapis; note that some cultic names like "Kelaeneus" are oddly half-Latinized).  Generally speaking you shouldn't give up on a name-search without trying a few wildcard-extenders, since there are some erroneous name-entries in Isegrim (*) and since the formats and spellings of Romanized Greek names will vary more often than not.  (Transcribed Greek forms are favored for Greek personal names -- but not place-names -- in Isegrim, e.g. Nikias, Herakles, Dionysos, Asklepios, but again this is not altogether consistent, as witness half-Latinized "Hephaestos.")  Similarly, if it's not absolutely clear who is depicted on either the obverse or reverse of your coin-type, you will need to review all the plausible candidates till you actually find a clear match; not all beards make a Zeus or a Herakles.  If you are not sure whether a left- or a right-facing portrait is historical or quasi-mythical, you should probably search vt:.* man l or vt:.* man r, since Isegrim reserves the term "portrait" for portraits of historical subjects; even a coin-profile of the personified Senate is classed as a "head" not a "portrait."  In connection with two-portrait coins where there's room for debate about which side is "heads" Isegrim sometimes wavers or duplicates entries; in those cases you should always search twice, switching entries for obverse (vt: / vs:) and reverse (rt: / rs:).

--- As I noted in the original posting, in Isegrim there is often an arbitrary side to what form an "exact" input takes (even place-names include some anomalies, e. g. "Trajanopolis" for "Traianopolis"); there are too many instances for me to note here, so you will want to note them for yourselves, for example, "pricecrown" for a prize-crown or urn, "temple-front of 4 columns" for the (four-columned) face of a temple, "stern decoration" for aplustre, "branch laurel"  or "wreath laurel" for a laurel branch or laurel wreath, and so on.  The most singular case I have noted: on coins picturing the Rape of Persephone, she is listed as Attribute rather than Type (a mere Attribute at her own rape?), and called "kore" instead of "persephone," thus rt:hades ra:kore, probably just because Hades is holding her.   Deployed various ways, there are other motifs that can show up as "types" or as "attributes," infants, animals, or birds, to name several; since there are a few verbs used to modify "animal" in Isegrim, i.e., fighting, jumping, sitting, or standing, you may do well to follow "animal" (and indeed other substantives, "wreath" or "branch" for example) with a wildcard-extender (= .*).  "Infant" oddly may call for a leading or preceding wildcard-extender ("vt:.* INFANT" or "rt:.* INFANT") to gain access to any tagged variants, for instance "HORUS INFANT" or "PLUTUS INFANT."

   (*) Some stealth-errors in the way names are entered in Isegrim produce entries that look right but actually aren't; I discuss one example at

Discussion topic 289816

If a type ought to show up in Isegrim, it generally will, but it may take a few variations on your search-parameters.

For a  census of coin-types that ought to be noted in Isegrim, but aren't, now see Isegrim Extended at

Isegrim| Extended| on the Classical| Numismatic| Discussion|

Not an easy example, and not very common (I've tracked down just three catalogued specimens of the type), with enough wear to help to explain the extreme misdescription on Xbax ("Pautalia -- Roma std. l. with Nike"); nonetheless with an Isegrim search you can get to a positive ID, complete with an image, searching only what's clear from the scan I've provided along with the details supplied in the heading (*) and the online materials they point you to.   If you're confident you've found the coin, please PM your ideas; at the start of next week I'll sum up how the entries unfold and post names of whoever's correct, unless you'd rather just stay anonymous.  Happy hunting for this one!

   (*) AE24 is the seller's description; the average diameter is 26mm.

I recognized the coin and was the underbidder; did you acquire it?  BM 66, but I won't say of what city!

The attributions of this major seller have been appalling for years now.  Quite often the only things right are the emperor and the metal; that info was used to call up some totally incorrect attribution from their substantial database, which was simply duplicated, the weight and diameter perhaps being corrected.


[0] Message Index

[#] Next page

Go to full version