I believe they made some electrically-driven machines earlier in the century, which used the "stepped reckoner" principle, or else the principle of a gear with individually-movable teeth that effectively could act as anything from a zero to a nine- (or ten)-tooth gear. (The teeth moved radially.)

Later in the company's history, probably somewhere between the late 1930s and 1945, Marchant brought out an electrically-driven machine with the same basic "carriage on top of a wedge" design, with rounded-corner and streamlined styling, colored in green and blue. Its internal mechanism was a dramatically-different, quite sophisticated, and more-complicated design compared to any others I know about.

It was over twice as fast as its competitors; so fast, in fact, that you could enter the digits of a multiplier serially on a row of keys along the right edge; even a nine was accumulated so fast that the time required was quite acceptable. I'm fairly sure its rated speed was 1,300 cycles per minute. Other electrically-driven machines ran typically at 600 cycles per minute.

This design was called "proportional gears"; each digit of the main keyboard had its own tiny 9-speed transmission (in the automotive sense); carries in the accumulator were done by 10:1 gear reductions, and each of the 20 accumulator dials (except the rightmost) had two differentials! (One added in the numbers from the little transmissions; the other realigned the dials.) That's a total of 39 differentials...

While accumulating a product of 2 or greater during multiplication, almost all parts of the machine were rotating at constant speeds; this was unique to Marchant. (The same happened if you held down the + bar; the relative quiet was quite impressive.)

By comparison, all other makes required the accumulator dials to start and stop, both for "digitation" and for carries. ("Digitation" is the process of updating the contents of the accumulator by transferring into it the number in the main keyboard's mechanism; it occurs before carries are entered into the accumulator.)

Marchant also made a calculator called the "Twin Marchant", which I know little about; there might even have been two quite-different special machines built. One account has it that two (earlier-design) Marchants were placed closer to and farther away from the operator, and possibly coupled mechanically somehow; the other possibility is that of a huge machine, possibly capable of multiplying two 20-digit numbers and displaying a 40-place product. The author would welcome more information on these! A good Web search is in order.

Marchant is also said to have made a (mechanical) so-called "binary-octal" (?) version; apparently this was a radix-8 (octal, or octonary) machine. Again, the author knows very little about this. It does sound practical, although the tooling costs would have been substantial.

Not long after Friden announced its EC-130 electronic desktop calculator, Marchant brought out its Cogito, which was similar to the Friden; the author's recollection is that it was physically smaller. It had peculiar-looking half-size zeros.

(Ernie Jorgensen has not yet seen this; he might well have some important additions and corrections, considering that he is a retired Marchant mechanical technician. -nb 97.10)