Actually I just tickled the X and Y values out of the graph to figure slope & intercept. (and therefore, forward cutoff voltage and equivalent series resistor value). Magic of EXCEL, and taking the two visually different segments

Vf = 0.51, Req = 7.3 ohm ... and then the second part in parallel:

Vf = 1.14, Req = 9.3 ohm ... in parallel with the first one.

HOWEVER, that being said ... there are a couple oddities in play: one is the forward voltage of 2.5 ... has a current-flow of 0.42 amps. My recollection of selenium-stack rectifiers was that each section (plate) is only good for like 20 volts reverse or something like that. Since they're technically a kind of Schottky, forward voltage drop can be low. How many plates though does this unit have?

As I found, there really isn't an exact silicon replacement. Pretty close is a standard silicon rectifier in series with a 7 ohm resistor, and from the diode-resistor point, another silicon rectifier with a 9 ohm resistor "across" the first resistor. Simple, and will have similar I:V characteristics. Just a bit higher shift in the forward conduction voltage, is all.

**GoatGuy**

[I researched CREE high-voltage Schottky SiC rectifiers... but they have a 1.5V drop or higher. Not much of a Schottky! And the silicon high-PIV diodes/rectifiers also seem to have the higher 1.4-1.8V forward voltage drop ... i guess 'cuz they put two lower-voltage rectifiers in series... (? anyone know ?) ... and there are no high-voltage germanium parts, and low-volt-drop conventional Schottky's also have depressingly low reverse breakdown voltage. So... I think the 'just use silicon' answer remains the best]