It didn't look too bad - Marchant XL

After fixing-up a Marchant Pony B,  was very curious for one of the early Marchant extending-segment mechanisms. The keyboard models are rather large (and shelf space is limited), so an XL (or XLA) was the machine to look out for. Ideally a 1920s specimen.

Several months ago, the opportunity arose to get a 1920s XL machine, the lever-set version of the extending segment range! It looked a project - thus relatively affordable - but likely complete and not too bad.

Or so I thought.

The machine was completely blocked solid - handle rusted in place, carriage stuck and the levers wouldn't budge. The counter clearing lever did rotate, but that was about all that did.

Taking the covers off, it became clear why very little on the machine moved.

The machine must have spent some time in a damp, cold environment - all the originally blackened parts of the mechanism are red with rust. And most are stuck solid. The rust goes deep.

This machine does deliver on experiencing a new, different design. Taking off the carriage on a calculator generally means finding and removing the carriage stop. On e.g. the Marchant Pony B this is obvious and easy, also on an Odhner it is usually pretty easy. Not here - it is hidden.

The carriage is prevented from sliding out by a tab on the front-left holder/rail that is downwards inside the cut-out area on the carriage base. Neither tab nor cut-out are visible when it is all assembled. To add to the experience, the rail is also pinned in place too. It needs to be lifted off straight-up vertically off these locating pins (and there isn't really space to do this vertically). The machine does deliver!

With rust this deep there's only one thing to do - tear it down and start again at the base :-)

That tearing down actually took several weeks too - with lots of oil, time to soak plus the occasional hammer-tap.

When apart, a first bit tackled was the carriage movement mechanism; also rusted solid into a 'block'. It is much more robust than the Pony's mechanism, will withstand the hammer-blows just fine. The spring-loaded latches (arrows) however are a tight fit in the sides - as if designed to get stuck solid with rust - and these need judicious, careful hammering to dislodge.

Another little 'boobytrap' is the registers. When one dial failes to clear it is not because of a sheared-off clearing pin as is usual, but simply a dirty/rusty spring-loaded ball - as was also noted by CvdV on the Twinplex (a double XL mechanism). Quite clever actually, and no risk of broken pins - but makes pulling out of the shaft a bit unnerving and unexpectedly hard.

The real surprise trick with the registers is that the carry-levers do not have a little spring-loaded rod as usual, but a spring loaded ball! Re-mounting a rod is tricky enough, but tiny spring-loaded balls is simply awful. Balls are nearly impossible to push down on their pointy springs and kept there whilst also sliding in the rod - the whole thing is as if designed to be a pea-shooter. (No photo - too busy chasing little ball-bearings.) Do not take out the intermediate-gears rod of these XL registers, if at all avoidable.

Many more parts to clean and make movable again. And then the challenge of re-assembling the whole machine will be the timing. The XL has a fascinating gear-train including counter-rotating gears on axles, doubling-back routing and no timing-marks anywhere that I could find (unlike e.g. a 1920s Monroe, with helpful marks on the exact teeth that need to mesh).

This XL will be a very slow and intermittent restoration project. A 'deep dive' into the Marchant XL mechanism as it all gets de-rusted and trying to re-assemble it all from the baseplate up.

Possibly it may work in the end, just possibly.

Remington Portable with approximate character

A regular, common-as-mud Remington Portable typewriter. It's a century old and of course still works.

It's the #2 variant Portable (as it would be called, after the #3 was introduced). This one had modifications done to the keyboard; the '3'  and the '7' were composited.

The '7' looks like a plus, just like the '4' already has. That is the label-split making it hard to soo, however. The top of the '7' typeslug has been ground flat and a thin sliver of a +/- charatcer soldered on. I.e. it got an around, approximate, just-about type!

The '3' had an equals-sign added, that is a bother to create on a typewriter otherwise so makes perfect sense. The machine however does still have an underscore and a plus, so a +/- could be typed with these two super-imposed (shift-4-6 and then let go of shift). A separate type is of course much neater and sits properly on the baseline.

Now what special use would require basic arithmetic characters as well as an approximate? 

In any case; the only machine I've so far with an approximate character!

Rabbit hole - toy tank and following the thread in online information

When looking for ads of the Baby Calculator, came across the Johnson Smith & Co company - sellers of novelty items. In a full-page 1930 ad, their toy tank description intrigued; what's the catch, how does this work?

The description claims it will move and climb obstacles, but "What makes it go is somewhat of a mystery, for there is no mechanism to wind up as is usually understood with mechanical toys,...".

In today's connected world with enormous amounts of information online, it was possible to find out about this! With the simple search for toy tank and "baby tank", some images from auctiuons over the past years were found to still be online. This shows the same toy, the below sold under the Animate Toy brand:

(Millers Auction Co - from a closed, ended auction)

The Johnson Smith & Co version more likely is a later, cheaper plain grey version; also still pictured at an antique store:

 
(Flying Tiger Antiques - from listing of item already sold)

This shows a picture of the drive mechanism underneath the toy - and a helpful patent date of June 20, 1916. With that date, the two patents that relate to this toy are easily found. These are US patents 1,187,659 and 1,187,660.

The first of these patents gives a very clear explation of the simple drive mechanism. The trick is to use a dampening material, instead of expensive gearing plus governor to slow down the spring unwinding. The single large driving mechanism is shown inside the housing of a toy bug.

The axle 2 is fitted firm in the housing, the flat (on the right) preventing its rotating. The fork 7 is fitted rigidly to the axle 2. The barrel 3 with wooden wheel 5 can rotate about this axle, and can wind and be driven by the spring. The barrel (tube 3 and end-caps 4) is filled with a molasses, pitch or tar-like substance that will sdampen and slow down the motion of the barrel against the fork 7. Very ingeneous and low cost.

The patents also give the name of the inventor; Edward S. Savage. Seeing the number of patents listing him as inventor, he was a prolific inventor - with about 50 patents, mainly for for toys.

   

Again the vast trove of online, scanned information yields more background; Mr Savage was featured in an illustrated article in an issue of an illustrated magazine of January 1939 - he even made the cover:

The full article can be read in the Mechanix Illustrated issue on The Archive. Pages 35 - 39, then 125 etc.

Mr Savage was a very active and also a very succesful inventor of toys - he's for example the inventor of the wobble-toys that are still made today.

At the bottom of page 39 of the Mechanix Illustrated magazine some toy bugs are pictured, those probably are the same toy that was drawn in the 1916 patent (applied for in 1914).

Again, specimens survive and get photographed when sold online, giving us color images. The pictures of one listing included a view of the bottom of the bugs, indeed showing the single drivewheel - one specimen perhaps lost its wooden wheel, or is a modified design with the end-caps (4) enlarged to also act as wheel. The cucumber bug still has its wheel, just as in the June 20, 1916 patent.

(Etsy - from listing of sold item)

So - the mystery of the drive is explained, and even revealed who created it. With portait and profile!

This little exploration, started by wondering about a line in the ad would have been hard to do a few decadeds ago, now easily possible through pretty amazing troves of online information!

Rabbit hole :-)

Baby Calculator go forth and multiply (a turd)

These are muddled times - with more information available than ever before, recent advancements in information processing have created something new - artificial information. Artificial information including falsehoods is available in larger quantities than ever before.

From a simple web-search for the Baby Calculator, one page really stood out in the search-engine results with its well-phrased explanation of the little gadget. A Symbol Of Social Progress - am intrigued; tell me more!:

Right away in the first paragraph it gets weird: "...one can identify a small thing called a “baby calculator,” which is a piece of equipment left from the period of the approach to reproductive health in modern society. This device, a rather humble tool by today’s measure, allows for a glimpse into the family planning strategies of the previous generations, ...".

One clear signal -apart from it being utter nonsense- is that the grammar sounds right at first reading, but the second half of the first sentence does not close. It seems something a statistical word-generation program would generate. I.e. this looks like artificial information from Artificial Intelligence (AI).

It gets better (or worse).

And it doubles down on this storyline.

The article ends with verbose platitudes that manage to still be surreal: "... the baby calculator of the 1920s is not just a forgotten relic, but rather a symbol of the humanity’s ongoing process of discovery and control over life. "

Normally I would not copy and show an article like this, but in this case did so to show/preserve it. It's an outstanding example of a moment in the development and application of AI. Pages like this are liable to vanish as quickly as they emerge. (But they do enter search engines' indices to muddle the waters, do poison the corpus of information.)

If wanting to have a look at the article, this is the link. The site by the way contains several other finely illustrated and phrased pages of articificial information. They're not all as hilariously wrong as the Baby Calculator page, but definitely all slop. And worryingly, slop with decent SEO.

This page and the whole site is created by a company Egeekowl that is based in Telavi, Georgia (თელავი, საქართველო), doing digital content creating for social media. The channel that this Baby Calculator pages is on/in is centred around a nostalgia theme. Likely with some form of advertising in or around it, this channel's content is claimed to reach an audience (~1M !), mainly in the US.

One of my kids reading the article reacted with derision and disbelief. Another, without paying attention to content, only noted the pretentious language as a red flag to then wonder "hm, is this AI?".

As signalled by Mr Polt's proposing of the term slopbait for a particular species of artificial information, we may need to develop a new vocabulary to describe the various types of AI slop that increasingly invades the human information-sphere. Would this be 'slopfeed'? Or 'feedslop'? Or slopbait! :)

Whilst this artificial information is silly and likely relatively harmless, it still is falsehoods being presented as bona-fide information. Does that make it not just silly, but also morally bad? And then considering the resources being spent (and burnt) for this, would that make it an evil? It is in any case not a positive or good thing, I'd argue. This is bad.

Coming back to the article; this is of course slop.

It's effective slop. Not great slop, only reasonably polished slop - perhaps a turd?

Baby Calculator and magic pricing

These are pretty common, sold in great numbers and many have survived. An American slide-adder of the Addiator or Troncet type. The Baby Calculator from Chicago:

This one is probably from the (late) 1930s and in very good condition. It may not have been used much, or at all. Even the box and the instruction sheet are fine and the original stylus is there too.

It was made after 1928 when the trademark was registered, and will not be later than March 1945. In April a new, updated design was introduced. Note that during the entire run of this type of Baby Calculator, the image in advertising was never updated and always said "patent applied for" instead of trademark registered.
 

The price clearly advertised on the box as $3.50 and again stated as $3.50 in the instruction sheet.

The odd thing is that in all advertisements found so far, the price is $2.50. It's $2.50 from the start of the Baby around 1925, right up to the new design first advertised in April 1945. Older specimens of the Baby Calculator do say on the box that the price is $2.50.

When was it changed to $3.50 ?  

Even more puzzling, although the company's own ads keep stating $2.50 throughout the 30s and 40s, it is also sold via American novelty mail-order firm Johnson Smith & Co for $1.50. And that is bundled together with a novelty magic notepad. Here listed in their 1936 catalogue.

To give an impression of the type of product the calculator is lumped in with; an example full-page magazine ad of the company from 1930:

So the price of this little calculator is a bit magical - somewhere between $1.50 and $3.50. Maybe the 3.50 number was printed to make a buyer feel better about paying 2.50.

As a slide adder the Baby Calculator is fine, but not great. Compared to an Addiator it is relatively flimsy and a bit rough to operate - frankly, inferior in quality.

It does work though. As with all slide adders, the claim of doing all four arithmetic functions is an exageration that might rightly disappoint a buyer. Perhaps even in the 1930s, this was more a novelty item than a serious calculating tool.

Giving the buyer the impression it was a dollar-off bargain may have helped :)

Very early LEGO typewriter - sighting

Spotted this weekend - an early LEGO typewriter.

When the company released set number 234 in 1958 with 1x1 bricks printed with letters, probably one of the things to make with these letter-bricks would have been a typewriter. The keyboard only has 21 keys (may have had a few more) and has an alphabetic layout. This model is relatively simple, but definitely a Lego typewriter.

This build was spotted in a display case in a museum / permanent exhibition / playroom for Lego. (This museum's been almost doubled in size since only a few years ago. It has all the recent themes and sets, and also a lot of history starting with the wooden toys, with loads of special and rare items. Very impressive.)

Applying some legology (by our resident legologist); the faded yellow, warped bricks suggest original 1950's or early 60's parts -that spent serious time in sunlight.

The grey baseplates with blank studs are correct for that period too, as are the round 1x1 old-style pieces. We've some period examples, the 2x4 a bit warped too:

The red 'platen' is made up from macaroni pieces - these do not have the stud notch cut-out to enable fitting on a baseplate so are the very early variant.

This early variant was made 1955-1956 - matching the 1958 letter-bricks. We've  only examples of the 1956-onward macaroni with the stud notch:

Transparent macaroni pieces with stud notch form the corners of the keyboard frame, they needed to fit on a base-plate. These were first introduced, transparent, in set 307 / 1307 of 1958.

The faded yellow bricks of the keyboard-base did show that letter-bricks were moved around, the outline of earlier positions are visible.

This early, 1960-ish, typewriter was in a display with mostly company promotion/sales material made for illustrating booklets or display by the trade. It is made from pieces that match a ~1958 build date. The use of yellow (uncommon back then) and a lot of macaroni also suggests access to ample supply of pieces (e.g. a company 'model room'). There always remains the possibility, however, that it was made later (or even very recently) from period parts, but likely this is a company created promotional piece. 

The museum -of course- also has the more recently issued LEGO typewriter on display:

Quite the contrast with the ~1960 version - this recent official typewriter is much more detailed - as is Lego as a whole, compared to the 1960 system.

Even though needing more imagination to 'see it', it definitely is a typewriter. An early one - perhaps this is even the very first official Lego typewriter! :-)

July 2026 safari - meagre

By special request, visited a large-ish charity second-hand store this week that's not too far away. The store sorts goods in departments, with the 'nice' stuff  getting its own department.

The typewriter shelf with the expected range of beige and plastic machines.

This Remington Noiseless Portable was the sole non-beige machine and kept in the 'nice stuff' corner.  An intact return-lever (often broken - a weak part!), but very obviously missing a spool-cover and the 3-key shows a broken linkage somewhere inside the machine. Play-damage, likely.

All left there - photo-safari only with no 'baggings'.