Scotch Modern might be worth looking at, though it's patterned on a face from later in the century IIRC.
Can you post an image to Typophile of the typeface you're trying to approximate? I bet good proposals of existing fonts would come from that which would be more usable as fonts than some scanned solution.

Thanks, Craig. Scotch Modern isn't half bad. I think, however, that my serifs are a little less Modern, a little more Transitional. Also, I think the stroke contrast on my capitals is a little higher.

A sample is now attached.

Take a look at Matthew Carter's Miller, which is a Scotch Roman but closer to the open forms of your sample.

Matthew Carter's Miller, although it has unbracketed serifs, is even more legible and well-proportioned than Century Expanded. This is good, for why bother to set out to make a font with flaws... but that means it doesn't closely resemble the typefaces of the bad old days.

So it might be that Hercules or Benton Modern Display are more like what he is looking for...

It's also possible that Scotch Modern or Scotch Modern Display, from ShinnType, would work; but Scotch Micro, like Miller, avoids the flaws of the old Scotch Romans. It's possible that the other versions do as well, and I'm not interpreting the appearance of the large specimens properly; the text specimen here

http://new.myfonts.com/fonts/shinn/scotch-modern/

in the sample text starting with "Recontextualizing" shows a beautiful, legible, and well-proportioned type - a good Scotch Roman instead of a bad one.

> why bother to set out to make a font with flaws...

Because every flaw is not sometimes?

hhp

Miller, although it has unbracketed serifs...

It doesn't. It has subtly bracketed serifs, which is one of the features that makes it easier on the eye than more extreme Romantic and Victorian types.

A common feature of many Victorian types is the tightly curled c and e, in which the bottom stroke curves upwards almost to touch the part of the letter above. Miller does not do this, and nor does the scan of the page provided by ‘plenipotentiary’. In fact, if you compare the proportions and treatment of letters in that sample with Miller, you will see that they are quite close; much more so than with Scotch Modern, which follows the tightly curled model.

Here is a comparison, showing what I mean. From top to bottom: the scanned book, Miller Roman, Scotch Modern:

Of course, Miller isn't a perfect match. Apart from inconsistent details such as the top of the t, Miller's heavier weight gives a different overall impression.

Another possibility is FB Escrow. Again, the details don't match, but they offer a light cut which might give the effect you're looking for.

Apart from inconsistent details such as the top of the t, Miller’s heavier weight gives a different overall impression.

So what you need is a Miller Light.

Sorry, horrible joke, carry on...

And to think, he came into this thread looking for a Scotch, neat! ;-)

Now, you guys are talkin'! :-)

ChrisL

Regarding the type, one type this reminds me of is Bell. It might be available in a digital form. I was fortunate to work at a press in Portland, The Anthoensen Press, years ago, and they had a few old original fonts of Bell. It is a very nice face of similar style. It might be suitable for your project.
Though the serifs are not quite as slab like as the original, I don't know, now I am thinking the Scotch might be better, or the other imaged above.
Scanning and using Fontographer is a possibility.
Aside from the fact that I am 2 years late, the Bell face is a little shown face with unique characters and I could only find this peck of a sample.
Maybe I'll dig up a better one just for the record. I wonder how this project turned out.

I like the Bell myself, especially the lc t. There is a sample of the Bell in Encyclopedia of Type Faces by Jaspert, Berry & Johnson,

Three ‘Scotch’ designs, by Matthew Carter:

Top to bottom: Georgia (1996), Miller Text (1997–2000), Monticello (2002). You may also wish to consider Vincent Text (1999), not shown here…

Scotch Roman is closest.
The vertical proportions are similar, and it has a round openness in the lower case, the ear of g is discrete, and the caps are much bolder.

Descenders shorter than caps would be hard to find :-)