Last Update: July 26, 2010 -- THE CURTA REFERENCE
+-------------------------------------------------------------------+ | This artice was translated by Andries de Man | | http://udel.edu/~ademan | | | | Copies of this article can be purchased from | | Charles Babbage Institute at University of Minnesota | | | | May 31, 2006 - Thanks to Jürgen Müller for finding an original | | copy of this article and producing much improved scans of the | | images. They're a big improvement over my 5th generation | | photocopy images! Also see an additional note from Jürgen at | | the end of this article about Backup magazine. | | -Rick- | +-------------------------------------------------------------------+
Peter Kradolfer
It started with a letter to Curt Herzstark in April 1988. I asked him if I could drop by, there were some things about the CURTA I was interested in. Only two days later the telephone rang "Yes, please come, I am in good health", which is not obvious at the age of 86. Surprised by such a spontaneity, I took the train to visit him. It was a day which I will not easily forget. "You shouldn't exhaust this man, he surely is of an age at which one needs some rest" was the warning I received from my wife when I left. I intended to return after three hours, at most. But that did not happen. After five hours of the liveliest conversation one can imagine Curt Herzstark stipulated laconically: "Well, we will never be able to finish, but please come back." I accepted this invitation thankfully, so which started as a single visit became a series of meetings. However, we never "finished". It is hard to tell the story of a rich life in only a few days. At first, Curt Herzstark impressed me with his technical achievements. The more I learned about his life, the more I was moved by the misfortunes he had and which he surmounted in his charming but tenacious way. I will tell about our conversations from a personal, i.e. not a historical, viewpoint. It will highlight many interesting details of his life.
He was born in 1902 as the son of a businessman in Vienna. Father Herzstark founded in 1905 in Vienna the company "Rechenmaschinenwerk AUSTRIA Herzstark & Co.", which built different varieties of the Thomas-calculators with additional patented Herzstark-inventions. The name "Austria Rechenmaschinen" [2] is well known among collectors of historical calculators. During the first world war the AUSTRIA factories had to make precision parts for shrapnel-fuses and collected by this occasion experiences in "Austauschbau" [3]. In 1916, after the "Realgymnasium" [3a], Curt started an apprenticeship as precision mechanic and toolmaker in the factory of his father. His mentor, Joh. Hayard, came from Glashütte [4], one of the centers of precision mechanics at that time. So Curt Herzstark was lucky to get a very decent education, despite of the war. Most important for his later occupation as an inventor of calculators was getting acquainted with the brand new method of "Austauschbau". After finishing his final test 'with honors' he started studying engineering at the "Höhere Staatsgewerbeschule" in Vienna, which can be compared with the modern swiss "Höhere Technische Lehranstalten" (HTL). Curt Herzstark did internships in his fathers' factory, where he worked in Assembly and Sales. In 1926 he was made responsible for reorganizing sales of Austria products in Czechoslovakia. Things could have proceeded steadily from there-on, but the tremors of world history made that impossible.
In March 1938, Austria was "brought back home" in the Reich by the Nazi's. As a consequence, the German army surveyed the industry of the annexed country [5], so they also examined the Austria-factories. The firm could save its existence by completing very difficult test-orders for the construction of gauges for the Army Supplies Office in Berlin. In other words: the company was not nationalized by the Nazi's, despite of the not purely arian [6] descent of Curt Herzstark, who had been promoted to technical manager in the mean while. In other words, the firm Austria was not nationalized, but had to adapt their production to the wishes of the Nazi's.
Everything went well, until 1943. Germany was already in its fifth year of war and failure after failure emerged: in May 1943 the capitulation of the German-Italian army in Africa, the fall of Mussolini in July, the advance of soviet troops at the eastern front and the American-British bombing offensives at targets in the North and West of Germany are only a few of the main events in this year. Curt Herzstark was arrested. He was accused of "helping Jews and subversive elements" and "indecent contacts with arian women". The only accusation they could hold against him was that employees, who were not at all inclined to racism, were caught at listening to broadcasts of the enemy (the English) and he tried to defend them. One of the accused was executed. When he was called as a witness in this case, Curt Herzstark didn't return home. Without any due process he was sent to Vienna in "preventive custody", as such an injustice was cynically called. After that he was put in a prison for Jews and finally deported to Prague, to the infamous Pankraz prison. Here he was taken over by the SS. Curt Herzstark told me things about the treatment in Prague that shocked me. I was very distressed by all the cruelties and the acts of contempt intelligent human being executed at that time. Curt Herzstark told me literally: "People tell a lot; I tell you only what I experienced myself", and he went on: "And I was even lucky, I went to Buchenwald" [7]. I startled, because I knew Buchenwald, and I knew that ten-thousands had died there. So in the fall of 1943 Herzstark was brought to the concentration camp Buchenwald "for special application". The reports of the army about the precision-production of the firm Austria and especially about the knowledge of Herzstark lead the Nazi's to treat Herzstark as an "intelligence-slave". Brought to the edge of desperation in the "small camp", he was allowed to move to the "large camp" at the end of 1943 and was called to work in the Gustloff factory [8] linked to the concentration camp. The name "Gustloff-Factories" was used for expropriated companies in Nazi Germany. There were many of those factories in Germany and they all worked for the military. The Gustloff factory of the Buchenwald used experienced workers that were transferred there from all over Europe. Generally speaking they were not Jewish. Soon Herzstark got special control tasks which he exercised - whenever possible - in the interests of his fellow prisoners; this put him at rather high personal risks. Apart from the manufacture of items for the military, there were also occasions of war loot repair. For example, after the retreat of the Germans from Italy in the summer of 1944, several truckloads of Olivetti manufacturing machines were transported to Buchenwald. Herzstark was ordered to prepare them for use. After that, the Thüringer manufacturers were invited to stock up with cheap production machines. Herzstark had to present them to the customers. He recognized one of them as the well-known "Waffen-Walther" [Arms-Walther], who was also famous for constructing of the "Walther Universal Calculator". [9]
In the winter of 1937/1938, after years of tinkering, the construction of the four-function calculator was essentially completed, and a first prototype was built and working (see Figure 4, Machine 1, at the far right). Two inventions - the complemented stepped drum and reduction gear - were filed for patent shortly after the Nazis took over power in Austria in 1938. Two patents were granted under DRP [11] No. 747073 and 747074, but no production was started; from the summer of 1938 on the AUSTRIA factories had to make measuring devices for the German army. On the other hand, Curt Herzstark didn't want to make his invention public for reasons of competition.
At first, his stay in the concentration camp Buchenwald seriously inflicted Herzstarks health. His condition improved after he could work in the Gustloff factories at the end of 1943. To his surprise the camp commander was already informed about his work on a calculator. He was ordered to make a drawing of the construction. They wanted to give the machine to the Fuhrer after the successful end of the war! That didn't work out. Anyway, Herzstark got access to a small drawing board and drafting machine and worked every spare minute, also on Sunday, at the drawings. Until the liberation in 1945 he had redrawn the complete construction from memory.
During one of my visits to Curt Herzstark I asked him about the origin of the
name CURTA. I knew that early drawings showed the name LILIPUT. He told the
following:
Liliput was indeed the original name, but the Society didn't like it. During the trade fair in Basel (1948) Miss Ramaker, trade correspondent of Contina AG said: "This machine is the daughter of Mister Herzstark. When the father is called CURT, the daughter has to be called CURTA." That's why the machine has been called CURTA ever since.
I went too fast. We have the idea and the name, but yet no machine. Curt Herzstark was still in Buchenwald. It was 1944, and the Allies threatened the German Reich more and more. On August 18, 1944, the Gustloff Factories were bombed, several hundreds of prisoners died and half the factory was destroyed. During the second bombardment a fellow prisoner lying next to him was deadly injured by a bomb fragment. Curt Herzstark survived miraculously. The part of the factory that was still operable was moved to a deserted salt mine in Billroda, 30 kilometers from camp Buchenwald. Here, 600 meters below the surface, they tried to resume production. Two days before the liberation the prisoners walked back to Buchenwald. Finally, the Americans liberated camp Buchenwald on April 11, 1945.
Curt Herzstark was alive and free. Additionally, he had a complete set of construction drawings for his miniature calculator in his pocket. Now the time had come to find a way to realize it. He contacted the Rheinmetall factories. And, as a kind of compensation, Herzstark was named director of the Rheinmetall factories and had to supervise the reconstruction. His fortune didn't last long, because in July 1945 Thüringen and Sachsen became part of the Soviet occupation zone, based on the Potsdam agreement. The few months until November 1945 were sufficient to revise the drawings and make three prototypes. These three prototypes still exist (see Figure 4).
The Soviets started to rebuild their parts of Germany according to their own plans. As a consequence, Curt Herzstark fled to Vienna, where he contacted a friend of the Herzstark family, the Swiss office machine manufacturer Jost. The take-over of production by Jost was almost established - imagine: CURTA, made in Switzerland - when the Principality of Liechtenstein showed up. Prince Franz Josef II of Liechtenstein [12] was trying to convert his poor farming country into a modern industrial state. And he succeeded, as one could read in the press at the anniversary of his government in 1988. The royal family invited Herzstark to build a factory for the production of his calculator in Mauren. It was a tempting offer, and so the Contina AG was founded in 1946. This development was a disappointment for the firm Jost, which was only partially compensated by getting sole representation for the CURTA. Curt Herzstark became technical director of the Contina AG. The real driving force and decision maker of the enterprise was a financing society. One could expect problems due to this difficult and obscure organization structure, and they emerged soon. At first, Curt Herzstark succeeded in manufacturing the first CURTA machines with a team of innovative expert mechanics [13]. The first production hall in 1947 was only provisionary, a ballroom in Hotel Hirschen in Mauren. At the same time production was started, a new factory building was being constructed for the Contina AG. One can easily compare the current building (Figure 5) with the factory from the 50's. A figure in the brochure "Calculation examples for the CURTA calculator", printed around 1955, shows on page 51 a picture of the "CONTINA FACTORY IN MAUREN; PRINCIPALITY OF LIECHTENSTEIN".
The first CURTA calculators were of a model with 8x6x11 positions; in 1954 model II with 11x8x15 positions [14] was added. The lack of technical expertise in the Financing Society, mentioned before, lead soon to a separation between the Contina AG and the inventor Herzstark. The AG was ended and its shares, so also the part owned by Herzstark, lost their value.
He was told that he could buy shares in a new AG that had to be established. But how could he pay ? Curt Herzstark didn't have any capital ! Herzstark was able to prevent the loss of his patents with the help of a distinguished Swiss patent lawyer. The patents were still on his name and not assigned to the Contina AG. Herzstark stepped back as technical director in 1951, and became a free-lance employee for some time. The end is easily sketched: The CONTINA AG was bought by the company Hilti in 1966, and the production of calculators was stopped in 1972. Until that date 80,000 CURTA I's and somewhat more than 61,000 CURTA II's had been made. I will explain the differences in construction in the second part.
So far about the story of Curt Herzstark, that was inseparably linked to the CURTA calculator. Many things of the life of Curt Herzstark fascinate me. One of them is the fact that I could speak personally with someone who had been in a concentration camp. For me, a Swiss who was 7 years old at the end of the war, this was an experience that made me experience a part of world history for the first time. But it is also the personality of the inventor. When I first visited Herzstark I was awed by my respect for him. I loosened up during later visits, and we became friends. On September 29, 1988, my last visit to Curt Herzstark, he said: "... you have to preserve my intellectual heritage...". Then there was his modesty. He told exciting stories about the old days, but he never put his achievements first. At first I believed that such an important person had to be wealthy. Wrong. Already during our first telephone conversation he said: "You will see I live modestly." That was true. Much later I tried to find out what was the cause of this. There seems to be only one: the fate of the inventor. The inventor has good ideas, but making money with them - that is done by others!
Peter Kradolfer [1]
This term needs an explanation of two points. "Complemented" indicates an algorithm that turns a subtraction into an addition. In other words: instead of a subtraction 481-247 the addition 481+x has to be performed (see text box). "Stepped drum" means a cylindrical driving element with protruding ribs of various lengths (see Figure 2).
The inventor Curt Herzstark has combined both ideas in a unique way in his CURTA calculator. The questions "What is the complement addition good for ?" and "How was it done in the CURTA ?" will be answered next.
How does one convert a subtraction into an addition ? The "complement addition" algorithm is based on the addition of the nines-complement of the subtrahend. The example in the text box explains this.
|
When I heard from this method for the first time, I was amazed that it worked. I have learned to appreciate the importance of the complement addition more and more - the subtraction in modern computers works in the same way. The only difference is that computers work in a binary system, so only with the set {0,1}. Mechanical calculators are based on the decimal system we are more comfortable with, having a set {0,1,2,3,4,5,6,7,8,9} of rank 10. During one of our conversations Curt Herzstark said about the complement addition of the CURTA: "The subtraction by the CURTA is kind of fake, it doesn't subtract at all, it only adds. The trick works, and that's what is important."
The method can be explained with a number circle but can also be proven mathematically. [5] In the following, we will focus on the technical realization of the idea.
"A calculator will automatically subtract by addition, when it doesn't transfer to the main counter the number set to subtract, but, instead, its nine's complement and subsequently adds a one to the lowest position. The construction problem is to enable a choice between two toothed structures to be used. The smallness of the available space and the expected lack of reliability prevented the use of movable teeth.The solution was a simple construction, consisting of a rigid body unifying two normal stepped drums, one of which is put upside-down."
Figure 2 shows the traditional stepped drum and Figure 3 and 4 the complemented stepped drum of the CURTA and its functioning. To understand the mode of operation, I once again cite Professor Holecek [2]:
"The one set of teeth, used for additions, starts, like any ordinary stepped drum, at the top with one tooth and ends at the bottom with nine teeth. These tooth segments are designated with 1 to 9 in the right part of Figure 4. Of course, their distance in a vertical direction is equal to the distance the setting gears travel when shifted one unit. The second group of teeth - the complement-Figure 5 [?] shows the details.toothing - is mixed in with the first in such a way that one and a half times the distance between segments below each segment of the normal toothing, there is a segment of the complement-toothing which adds up with the first one to nine. This system allows one to subtract the set number by simply pulling up, using the crank, the stepped drum over one and a half times the distance between segments, and then turning it in the same direction as for addition. During this operation the 11-digit nine's complement is automatically added, and the first position is increased by one."
Holecek continues: "The complement segments are not in operation during addition, because they move between the possible positions of the setting gears. When the stepped drum is shifted up over one and a half times the distance between the segments (Figure 4, center), then the complemented toothing engages with the setting gears, and each gear of the setting mechanism does not rotate over the number of teeth corresponding to the number set, but over its nine's complement. In Figure 4 one can see this for the number 971853. The center of the Figure shows the stepped drum in its position for subtraction. The toothed segments of the complemented toothing are designated with 1' to 9'.
In order to transfer the complete 11-digit nine's complement for the subtraction, nine has to be added to each of the three foremost positions, since they do not engage with the setting mechanism. For this purpose the three foremost shafts have fixed gears in the zero position, that are engaged by the 9 tooth segment 9' in case of subtraction. This way the 11-digit nine's complement is truly added if the stepped drum is in its upper position."
As shown in the box, subtraction needs, apart from adding the nine's complement, also an increment of the lowest position with 1. For this purpose, the shaft at the first position is provided with a second setting gear. Before we go into details, we would like to discuss which advantages the described construction has.
133x89 = ?
Someone reproducing this algorithm for the first time might find it rather complicated. It would be much easier to crank until ... well, until you make an error. The more I crank, the sooner I turn one revolution too many or too less. Of course one can correct this without any problem. But one has to see the error first. In practice, I just crank until the number 89 shows up in the revolution counter (see Figure 1), without having to think if one needs an additive or subtractive revolution - I watch the revolution counter and always switch modes at the right time. This is an essential difference between the operation of a CURTA and any other hand-driven calculator. Before I wrote down this sentence, I tried to do the same operation - 133x89 - on one of my other mechanical calculators, using the method of progressing complement addition. With my TIM [7] I can also, instead of 133x89, calculate 133x(-11+100), the result is the same. The big difference is that, after the calculation, the revolution counter shows 111 instead of 89. The numbers of the tens and units are red, but is not of that much use to get the right multiplicand 89. This is especially important in chain multiplications, like for instance 38x14x67x63.44=?
During the production era of the CURTA, i.e. from 1947 to 1970 [8] CURTA's were accompanied by a little pamphlet "Rechenbeispiele für die CURTA" [Examples for the CURTA]. It contained many other examples. It is an exciting passtime for any modern amateur of mechanical calculators to (re)discover the finesses of its practical application. The pamphlet shows practical applications in the fields of trade and industry, statistics, engineering and geodesy. Geodesy or surveying has always been an important application field for various methods of calculation. Alas, I can't go into details about this, I only mention the calculation of arithmetical averages and standard deviations.
Let's look at the original patent applications. Or better, the two mentioned before, because there are more than 30 (!) patent applications in 14 countries in the name of Curt Herzstark that have something to do with the CURTA. So I will limit myself.
The first one, DRP 747073, valid since August 19, 1938, reads under the
heading "Claims":
"Calculator with one single stepped drum and a mechanism for performing
subtractions with changing the direction of revolution of the stepped drum,
characterized by establishing onto the stepped drum, apart from the normal
stepped toothing, a second rigid complemented toothing, in such a way that by
an axial displacement of the stepped drum the normal toothing or the
complemented toothing can be brought into engagement with the setting gears."
This long sentence, which summarizes the claims very well, is followed by
three sections completing and detailing on the claims. So we have seen that
the first patent relates to the complemented stepped drum, without
mentioning this term explicitly. The second patent, like the first one granted
in Vienna, but designated DRP [9] and "decorated" with
the German eagle and a swastika due to the changed political conditions,
protects the transmission gears.
I cite the first line of the claims: "Calculator with only one stepped drum and with setting and counting members arranged in a circle around it, characterized this way, that the transmission mechanism between the setting gears (4) and numeral dials (13) has a gear (10,11), which extends the time of engagement of the stepped drum with the setting gears, thus distributing the friction of the counter members over an as large as possible movement of the crank and reducing the velocity of the numeral dials". End of citation. It is followed by three more points. The numbers in brackets refer to the drawing shown in Figure 7.
Figure 8 shows parts of the gear in a cut-away model. Because I didn't fully understand the cited line of the patent, I studied some more papers to find out the idea of the gear. Studying the open CURTA, the cut-away model and conversations with the constructors finally helped me to understand the idea.
The essence is that the setting gears on the driving members and the crown
gears at their tops have five teeth (see Figure 8 and 9). On the other hand,
the gears on the numeral dials have ten teeth. This unveils what the cited
patent meant by "extends the time of engagement of the stepped drum with the
setting gears". To transfer a nine, the driving member, which consists of a
shaft, a setting gear, a tens-
As already said, apart from the two main German patents, there were many
additional patents in Germany and all over the world. As an example for all of
them, I give the patent claim of the significantly younger swiss patent from
1950: "Calculator for all four arithmetical functions with only one stepped
drum and setting and counting members arranged in a circle around it,
characterized by having setting and tens carrying gears with five teeth
each and having a transmission between setting gears and numeral dials with
a gear in such a way that the numeral dials make a tenth of a complete
revolution per tooth of the stepped drum." It continues with 9 (!) additional
claims.
For the CURTA, and I will restrict myself to the CURTA I, it is really
difficult to find constructive differences. However, there are some.
I will comment on changes in the case, the setting knobs,
the zeroizing lever and the crank.
"When screwing on the top it was possible that the crank would also turn a
little bit to the right. This caused the machine to be not ready for use.
There were several locks that prevented it. To release these locks one had to
turn the crank until it came back to its resting position. This was not only
laborious but also annoying. The left-turning thread prevent this source of
errors - the crank cannot be turned left anyway." Indeed, a small turn of the
crank is enough to block further operation of the machine. The real purpose of
the locks engaging in such a manipulation is to prevent operator errors.
Several locking mechanisms had already been protected by patents in an early
stage, showing their significance. I won't go into details here.
Curt Herzstark told me another nice story about the case. During a visit to
the CONTINA AG, Prince Heinrich of Liechtenstein brought a leather case he
had been letting made for the CURTA. "Such a delicate machine has to be
cuddled snugly, the tin case is much too rigid.", he said. Director Herzstark
wanted to see the thing more closely. When handing it over, the machine almost
fell to the ground because the leather top cover became loose. This took care
of that new idea. During some time the tin cases were replaced by plastic ones.
According to Curt Herzstark they didn't meet the requirements, so one returned
to the tin cases. That's why nowadays it is the final aim of any calculator
collector the possess a CURTA in a tin case with a normal thread !
I wanted to know from Mister Herzstark why they changed the setting knobs.
He told me: "Originally I believed one should be able to see the numbers to
which the slides were moved. That called for a setting knob that was as thin
as possible, so it would cover the numbers as least as possible. Later we
found out, one'd better concentrate on the numeral dials. This allowed for a
more finger-friendly design of the setting knobs." [12b]
Which difficult constructive problems had to be solved, and how were they
solved?
Künzli: "One problem I worked on for a long time was the construction of
the stepped drum (final design see Figure 3). First I tried to mill and
lathe some out of a aluminum block. Although I succeeded, I had to agree
that the method was much too elaborate to be used in mass production. After
some brainstorming the idea emerged to use die-cut parts (Figure 15).
This way we could use abrasion-resistant steel. The only problem was the
tolerances. The stepped drum consists of 37
segments [15] - a positive
tolerance of only a hundredth of a millimeter already leads to a stall.
For subtractions the stepped drum has to be pulled up only 3 mm. The mentioned
positive tolerance adds up to one-eighth of this displacement - so it couldn't
work. For a first series of 6 prototypes the segments had to be ground to
their prescribed size. Even then, only 2 machines of this series worked.
The second series was better - 4 out of 6 worked. Later we found the solution.
Each segment had to be bent a tiny bit and have a negative tolerance.
By pressing the segments together more or less we could force the stepped drum
to its prescribed size."
Kessler: "It was extremely difficult to mill the spiral groove in the
setting shafts (Figure 17). The tools we used wore extremely fast.
There were no tools out of hardened steel for these small dimensions.
Not only the wear of the tools bothered us, also the surface roughness of the
groove. With a special drilling oil we succeeded finally to make grooves with
the right roughness. When in use, the movement of the sliders had to be
smooth, without jumps. Lubrication was out of the question. Tests showed that
dust and oil soon formed a kind of resin. Finally we found a solution that
didn't require any maintenance - the little guiding screw had to made of
bronze. This ensured smooth sliding over the steel surface. Similar problems
with wear showed up for the reset cam of the tens-bell (Figure 18).
After several trials we finally found the right chrome plating."
Why were the numbers engraved in the numeral dials, and not printed ?
Künzli: "The numeral dials were anodized black. The numbers had to be
clearly visible, that's why there were engraved after the surface treatment.
The anodization posed a typical problem. In the beginning, the layer
thickness was too large, which resulted in problems during operation."
You have told that the counter ring was cut and milled. What was the
reason for this elaborate method ? Were there no simpler and cheaper methods ?
Kessler: "Indeed, those methods existed. But this part had to pushed up and
rotated a decade for each multi-digit calculation. This required that the ring
shouldn't slip in any case, even after years of use. The simple process of
"randrieren" [15a] was out of the question, it
couldn't be done deeply enough."
Which is the the nicest experience you remember related to the development of
the CURTA ?
Kessler: "The service manual that I could produce. And the first
presentation of the CURTA at the Basel Trade Fair in 1949 (see Figure 21).
It was very exciting to share my enthusiasm for the machine with the people
crowding around the booth. There were many professors, surveyors and teachers
who wanted to know every detail - though the teachers usually didn't buy
anything."
Künzli: "Yeah, selling at trade fairs, we enjoyed that. Especially in the
beginning it was not a thing we could count on. There were some dark periods."
Please tell !
Künzli continued: "For instance the rejection of a very large American order.
At the Basel Trade Fair in 1949 an American came to the booth and wanted to
order 10000 CURTA's [16] for retail in American department stores. This went
beyond my competence so I called the manager of the Financing Society. He simply
rejected, saying that the CURTA could only be sold in specialized shops."
The Principality of Liechtenstein didn't have any qualified precision mechanics
after the second world war. So the CONTINA AG tried to attract mechanics by
advertisements in the swiss daily journals. Mister Künzli responded to one of
those advertisements in the Neue Züricher Zeitung. During his job interview
he was told he could also bring his wife because the firm needed a secretary.
That's how Hans and Paula Künzli became two of the four first employees of the
CONTINA AG and made a significant contribution during the first pioneering
years. Mrs. Künzli, which events from this pioneering time do you remember ?
Mrs. Künzli: "The start at April 1, 1947 with wild plans and the contrasting
reality: production started in the ballroom of the Hirschen Hotel in Mauren,
there were only the four of us. After that a nice period started: the expansion
until 1951, when a conflict arose between the Financing Society and Curt
Herzstark. It was around Eastern 1951. One fine morning the postmaster told he
was not allowed to hand me the mail. Mister Herzstark was denied access to the
factory. It came as a cold shower for me. I already had had the impression that
the Financing Society would give our director Curt Herzstark more troubles
than support. But I didn't expect they would go that far."
An explanation of this incident.
From early on, the Financing Society had financial problems, for obvious
reasons. By selling new shares the CONTINA AG had to be put to a sound
financial basis. This meant that all old shares completely lost their value.
Also the significant part owned by Curt Herzstark. Obviously, he didn't agree,
which lead to the conflict mentioned above. Aided by a eminent patent attorney
from Zürich the loss of patent rights could be prevented. Despite of that,
Curt Herzstark worked only a few more years for the company as a free-lance
consultant.
My interview had not come to an end yet. We talked extensively about the
relation between the people of Liechtenstein and the swiss craftsmen. The
relation was not good, on the contrary. In a public speech someone demanded
"... to educate our people as fast and thorough as possible ... so we can get
rid of the Swiss." On the other hand, the relation between the employees and
technical director Herzstark was superb. I remember the words with which
Curt Herzstark gave me the addresses of the interviewees: "Arnold Kessler and
Hans and Paula Künzli, you have to talk with them. Without their help the
CURTA could not have become what it is" and he added laconically, "they were
great guys !"
I don't know if I succeeded in completing the task Curt Herzstark gave me "to
preserve my intellectual heritage". Surely I will never forget my encounter
with the CURTA, its inventor and its constructors. I hope these two
publications entice the reader not to forget the CURTA either. I also made some
new contacts. These and the many responses are both a reward and an incentive.
An incentive to continue my involvement with historical calculators. Maybe you,
dear reader, can help me with that. I'm looking forward to your response.
Note: From Jürgen Müller: When I first came across the article on your web site, I was curious to
find out more about "Backup" magazine, hoping it might contain more
information on historical calculators. It was hard to track down,
considering the very common title word... I found a copy at local teacher's library.
Turned out that "Backup" was aimed at high school teachers of computer science.
It was mostly concerned with fundamental algorithms and programs for then-modern microcomputers,
but had a few articles on the history of computing (often by Peter
Kradolfer). Hence, it's not too significant a resource for calculator
fans. Nevertheless, here's the full bibliographic data I found:
Title: Backup -- Informatikzeitschrift für Schule und Weiterbildung.
Publisher: Sauerländer Verlag, Aarau, Switzerland
Available issues: 1.1986 - 5.1990
ISSN: 0258-4891
I did have some second thoughts about copyright. While the "Backup"
magazine has faltered in 1990, after just a few years of publication.
Changes in construction details of the CURTA
I already mentioned in the first part of my
contribution [10] that the CURTA
hasn't been modified much of a few decades. All this time (1947 to 1970) there
were only two models, the 11-digit CURTA I and the 15-digit CURTA II. In
contrast, the most famous German calculator, the MERCEDES-EUKLID, was put at
the market in 21 (!) models. [11]
Case
For the early models the top of the case was screwed onto the base with a
normal, right-turning thread. For later models there is a left-turning thread.
I asked Curt Herzstark why it was changed. Here is his answer:
Figure 12 and 13:Counter ring with setting shafts and the setting
numeral dials connected to them. Clearly seen are the two versions of the
setting knobs, before and after the end of 1948, and the identical way they
run in the spiral groove.
Setting knobs
In the patent drawings (see Figure 7, backup 1/89) show cylindrical pins
attached to the setting slides. In the early production they were made this
way. Later [12] the shape of the setting knobs was
changed, the difference is
shown in Figure 12 and 13. When looking at both pictures one sees that the
bearing of the setting slides has not changed, although there has been 15 years
between the two machines shown. A screw-shaped groove, in which glide a bronze
guiding screw and a steel ball, transforms a vertical movement in a rotation.
The speed [12a] of the groove is exactly ten times as big as the distance
traveled by the setting slide when displaced over one unit. Hence such a
displacement turns the numeral dial that is connected to the setting shaft to
the next number. The balls have a diameter of only 2 mm and are pressed by
springs into the cavities that are visible in the pictures.
This stops the slide at each number and allows for accurate setting (see also
Figure 17).
Zeroizing lever
It occurred often that the ring of the zeroizing lever (see Figure 14) broke
off. Because the machine is round it could easily roll off the desk when not
used carefully. Originally, the machine had to be disassembled to replace the
lever. To avoid this, the zeroizing lever was reconstructed in such a way that
it could put mounted on the clearing plate from the outside, without having to
take the machine apart.
Figure 14: At the left the early machine nr 4086 and at the right the
later machine nr 63122 with their differently shaped cranks. The ring-shaped
lever under the crank is the zeroizing lever. Below that is the clearly visible
counter ring, which has to be pushed up and rotated one position during
multi-digit calculations. To obtain a deep ribbing that doesn't slip the
ring was cut and milled.
Crank
An exterior change was made for the crank. Originally it was round, but later
on it got edges. Figure 14 shows the difference. It simplified the
manufacturing and reduced costs. As I already told in a previous contribution,
the CURTA cost about 450 Swiss francs in 1950. Converted to todays' currency
it would have cost 1500 Swiss francs [13] a price
which is higher than what
collectors pay for it now. I expect that the collectors' price will increase,
CURTA's become rarer and more sought after. In the 1950's, its price was
really high for a four-function calculator, but not without reason.
Manufacturing it was very complicated. For instance, a lot of the numbers were
engraved. The counter ring was cut and milled. Undoubtably, the advantage
of this elaborate production was an extremely high quality of the finished
product. Indeed, even now it is hard to find a CURTA that doesn't work.
This is not very common for precision equipment that has been standing still
for decades. Any collector can tell you stories about it. Often enough he has
experienced some anxious moments when cleaning a machine - everything blocked !
Figure 15: Some parts of the CURTA. In total the CURTA I consists of
687 parts, 139 of which are different. At the right the star of springs,
used to prevent the counters from turning to far. Top left a setting shaft
with setting slide and fork for shifting the setting gear. Below four segments
of the complemented stepped drum for the numbers 1 to 4. Between these segments
and the star of springs a numeral dial. The photographed scale indicates the
size: each block is 1 centimeter.
Figure 16: The two bearings shown are part of the carriage and bear
the main shaft. The label with the tolerance specification illustrates the high
accuracy needed for manufacturing parts of the CURTA.
An interview with CURTA constructors
My acquaintance with Curt Herzstark also brought me into contact with the
constructors of "the early days". I met the mechanics Arnold Kessler and
Hans Künzli [14] and could ask them some questions.
Figure 17: Setting slide with setting shaft. The spiral groove is shown,
that guides the little bronze screw, as well as the shallow cavities the rest
the small ball. The slide can be halted at each position corresponding to a
number.
Figure 18: Tens-bell with at the very bottom right the reset cam. Left
from it a carrying lever made from chrome plated steel. A pin on the
numeral dials shifts down the carrying lever. When the tens-bell, which is
rigidly connected to the crank, is rotated, it pushes back the carrying levers
and the carrying gears they operate (see also Figure 24 to 26)
Figure 19: A opened and a cut-away model, both with the finger-friendly
setting slides. At the right of both machines the reversing knob is visible.
Sliding it down makes the revolution counter count additive revolutions
negatively. This can be useful in a lot of calculations.
Figure 20: The manufacturing building of the CONTINA AG in Mauren,
Principality of Liechtenstein, to where the company moved in 1948.
Picture from 1952.
Figure 21: The first presentation of the CURTA at the Basel Trade Fair
in 1949. In the center the inventor Curt Herzstark, at the very left
Hans Künzli. The other gentlemen are (left to right):
von Gerlicy, Asal and Maier.
Figure 22: A "sample suitcase" that was used when visiting customers to
explain technical details. It contains: an open model, below which a
carriage without clearing plate, above the name tag a complemented stepped
drum and at the very top a lower main casting with its many holes to
accommodate the transmission and setting shafts. One can find it at the
bottom of a complete machine. At the right there are two setting slides,
two transmission shafts and two main shafts. At the right bottom the so
called machine body.
Figure 23: Manufacturing room of the CONTINA AG in 1951, Curt Herzstark
is making an inspection tour (standing at the rear left).
Figure 24: A partially disassembled CURTA I. Exposed are some
transmission shafts and two carrying levers, that are operated by the numeral
dials above them and slide the tens-carrying gear down within reach of the
tens-carrying tooth, that lies above the revolution counter toothing, but just
can not be seen in the picture. On the other transmission members one cane see
setting gears.
Figure 25: Disassembled carriage seen from below (left) and the machine
body (right) which is mounted below the carriage in the complete machine.
Two tens-carrying levers can be seen.
Figure 26: The same machine body as in Figure 25, photographed from
below. A tens-carrying gear, that can be shifted along the transmission
member, is shown slid down.
Figure 27: The "business card" of the CURTA, the bottom of the machine
with the serial number, from which the collector can deduce the year of
production.
The tens-carrying mechanism of the CURTA
The description of this ingenious construction is a must. Space limitations
restrict me here to showing only a few pictures. I hope Figure 18 and 24 to 26
show enough details to clarify at least the principal idea to the interested
reader. Maybe its description will be given in another contribution, or a future
book. Maybe another reader will pick up the thread and publish his own
contribution.
Literature used:
Notes:
The Calculator Reference by Rick Furr (rfurr@vcalc.net)
Back to The CURTA Calculator Page
Back to The Calculator Reference