Reference points for variety descriptions.
RogerB
Posts: 8,852 ✭✭✭✭✭
For those working on updating or revising variety attribution guides, the following suggestion might produce a long-term improvement.
Select a set of clearly defined reference points on the central device – portrait, wreath, star, etc. – and make all measurements relative to only those points. This is done occasionally with variety descriptions for some series, but much confusion could be avoided if it were applied consistently.
The purpose is to avoid defining thinks like date or mintmark positions in relation to features that are not fixed, such as denticles or stars. (One of the major confusing factors for new Morgan dollar variety collectors is that date positions are described in relation to denticles, yet denticles are not fixed with respect to the portrait. In this instance, use of the front of the neck as a reference will produce more easily measured and communicated date positions.)
Just a small suggestion.
Comments
@RogerB...Good suggestion, and what you have outlined is a good step towards real standards in coin assessment. Standards are definitely possible... and eventually inevitable. Cheers, RickO
Thanks! Using this approach will also enable computer assisted variety identification - even for obscure or difficult to see pick-up-points. A person should still verify the system's identification, but it will vastly increase accuracy and reduce in-house processing time. For coins with thousands of known varieties and potentially many others (such as VAMs) this would greatly simplify identification.
(photo courtesy of PCGS)
I do this for Liberty Seated Half Dimes.
For 1840-1873 "with drapery" I draw a vertical line down from the skirt pendant/bulla and note where it intersects the date digits. For 1837-1840 "without drapery" I use a similar location between 2 skirt folds.
The dentils are on the hub for this series, though - so this line intersects the dentils in a predictable location.
There were a few hubs used, so this location varies a little bit over the years, and sometimes a year has 2 hubs.
Note: to be able to draw a vertical line in a reproducible way, you need 2 things:
1. A precise and easy way to rotate the coin photo so that it is "straight".
For the 1840-1859 "with drapery" hubs (there are 3), I rotate the photo so that the centers of stars 2 and 12 are level.
The Irvanview software (free on Windows) has an easy command to make 2 points level.
2. A precise and simple way to draw a vertical line on a bitmap grid.
The Irvanview software does this by holding down Shift when drawing a line.
For more details, see:
https://web.stanford.edu/~clint/hdag/half_dime_attribution_methods.pdf
The suggestion is to select two (or more) points on the central design, then connect them with a line. This line becomes a stable reference for all pieces of that design regardless of hub changes, rotation or anything else. If these points are part of both w/ and w/o drapery then use the same for all coins.
In the 1847 example, The "vertical" line is arbitrary and thus unstable. If one bisected the junction of skirt pendant and connected that to another fixed part of the figure, then a true reference would be created.
Every other part of the design is measured in relation to the central reference, only.
Don't know about anyone else....but I've never put a ruler on a coin, and "measured" anything while doing attributions.
If they tell me something is "2.1mm from" something else, I ignore them, and look at the photograph.
As for the 1847 above, it seems to me that for THAT variety, which by definition comes from a single die, that line IS perfectly defined. If your coin doesn't have the positioning shown in that photo, then it's not that variety!
The fallacy is the "vertical" line does not connect two stable points. A stable reference system does not change with the variety; if it did, then it would be useless.
As for measurements, for those to whom such specifications are significant, there are much more accurate and safe ways of measuring that putting a ruler on a coin.
The suggestion is not that any existing variety description is "wrong," simply a better, stable system can convey much more information with the same, or less effort.
@RogerB,
I had forgotten to include some important details with regards to drawing that vertical line, so I added them to my post just now.
I think you will agree that with the full method described, it is a stable (and reproducible) system.
The linked PDF describes all my chosen location metrics (both date and mintmark) for all hubs used in Liberty Seated Half Dimes.
Better. But using stars as reference features for rotation doesn't really work. Stars were added in relation to the central figure, not the other way around. Therefore, the reference points must be part of the central figure and only that device.
The stars are on the hubs for the 1840-59 "with drapery", so they are in fixed position with respect to the central figure, for a given hub.
You are correct that the stars 2 and 12 are not as useful for 1837-40 "no drapery", because they are not on the hub (were added by hand). I used 2 points on the central figure for that hub.
An added advantage of stars 2 and 12 is that they are quite far apart, so they make a rather precise rotation.
(Any small errors in locating the star centers does not change the line angle as much as it would for target points closer together).
And they are visible down to very low grades.
Using points on the central figure does not work when the central figure changes, as it did for half dimes in 1840, 1857, 1859 (and maybe 1860; I don't have my sources with me at the moment).
In my initial efforts, I tried using a line connecting the back of Liberty's neck to the skirt pendant / bulla.
But I found the edge between the back of Liberty's neck and the field was often obscured by shadow due to the light angle.
So I like using the center of the pendant, instead of the edge of something, for the same reason.
If the same points are on each version of the central design, then they will work for all. If not, then new points would have to be defined. You cannot use stars, denticles, inscriptions, date or anything else. The master dies were ALWAYS prepared with the central motif first.
I think my point is that any markers that are on the hub are suitable as fixed points of reference for comparing dies made with that hub.
Starting with the 1840 "with drapery" half dimes, everything was on the hub except the date and mint mark.
So the stars and dentils all work.
On some of the 1853-55 with arrows dies, the date and arrows are on the hub as well.
The main trick is making sure you are comparing dies from the same hub.
On series where there are several different hubs used per year, the above method would not be helpful.
But for LS Half Dimes, there are very few hubs across the whole series.