I wonder whether this is not a v, but a lowercase Greek nu? It could be that the original page was set from a Greek or mathematics font, not from from a Latin font. That would explain the different f, too. Maybe the distinctions were lost over time.

I found a Wikipedia page for nu, and a a Unicode specification.

--John

i believe Cochin has this form of v.

I wonder whether this is not a v, but a lowercase Greek nu?

That's a reasonable guess, but the texts do show show v, u, nu, and upsilon and they are quite distinct. The nu looks like the sharp v, but the left arm is angled out instead of in.

Cochin is interesting. According to Linotype's blurb, "The Paris foundry Deberny & Peignot was the first to release this design in 1912...This version was created in 1977 by Matthew Carter for Linotype."

The typface you see in the picture was used all through the 1800s. So, Deberny & Peignot created a modification of this 19th century face, then Carter created a 1970s interpretation of that. Carter's font is definitely for display.

This is my analysis of the images:

The basic italic v in this typeface is the rounded form. This fits with the conventions of the period and the style.

The pointed form is a variant specifically for use in setting mathematics, where it can be used without fear of confusion with the Greek upsilon.

The use of the pointed v in the word 'positives' looks to me like an accident, although I suppose if it is used consistently in this way then some other pattern is at work.

Cochin is interesting. According to Linotype’s blurb, “The Paris foundry Deberny & Peignot was the first to release this design in 1912…This version was created in 1977 by Matthew Carter for Linotype.”

Try contacting Matthew Carter and asking him why he added it to the 1977 Cochin.

After much further delving in the stacks, I've found at least part of the info I was looking for. The usage described above was the house style of Gauthier - Villars.

They employed this style from at least 1864 to 1910 and they accounted for the majority of math texts published in France during that period.

You can browse online scanned books by Boussinesq, Cahen, Lacroix, and Laurent.

[Edit: I have also found this usage in a Greek/French edition of Thucydides' History of the Peloponnesian War published by Firmin Didot in 1868.]

Acanthus is very reminiscent of the body text in Boussinesq, at least. Acanthus, of course, is not a math face.