English Translation: Theoretical and Practical - Golden Chain of Homer by Naxagoras

[Andro]

Especially on the toilet. This area is particularly dangerous and prone to lightning strikes.

However, Elvis died on the toilet. So did Catherine the Great, Judy Garland, Tywin Lannister and others

 

[JDP]

Ok. I wanted to give you a summary of the first part of the book, but the introduction told me not to say anything to the unworthy, or I and my soul are going to be cursed in the name of god and I must expect an early death.

Now I'm afraid a thunderbolt will strike me while sitting on the toilet. Do you think that's exaggerated?

Watch this episode of Rod Serling's Night Gallery very carefully:

[video link broken]

 

[Christophorus]

The translation of it would be something like "Naxagorae's experiment, following Plato's rings and Homer's golden chain.
Does anyone know an EN, FR, IT or ES translation of it?

Probably could be of some use to this study also.

Check it here:

Johann Ehrd von Naxagoras - Experientia Naxagorae (German), 1723

Ganzer Titel: Experientia Naxagorae, secundum Annulos platonicos, et Catenam auream homeri. Worinnen der wahrhaffte Process, die Universal-Medicin zu elaboriren, so wohl vor den menschlichen Leib, als die Metalla zu verbessern; klar und aufrichtig vor Augen lieget. Johann Ehrd von Naxagoras ist...

issuu.com

 

[Illen A. Cluf]

totally with you - if you organize payment. I have done my deeds and more re: crowdfunding, but best of luck.

I doubt it though. considering the echo on my earlier efforts towards such an undertaking.

20 people would likely not be enough. nobody pays close to 100 bucks for a book like that. try 50 people and you may have some success.

also consider that the cost I suggested are meant for the rights to stay with the translator - the crowdfunders would get a relatively costly copy and everybody else a much cheaper copy later when the book is published thru LULU or some such.
So you are looking for 50 generous souls.

-sam

I know that many here would fully agree that Aurea Catena Homeri (Golden Chain of Homer) is considered to be one of the most descriptive and least ambiguous manuscripts of alchemical lore. It is more focused on outright experimentation, and covers the processes of fixation, putrefaction, and generation in far more depth than most works of the era. It moves away from the more destructive alchemical approaches involving such ingredients as Antimony, and seeks a far more natural process of achieving the Philosophers Stone that involves man’s urine. The Golden Chain is similar to the approach used by other Alchemists, such as Goosen Van Vreeswyk.

The Treatise mentioned in this thread seems even more valuable, as it actually provides a very serious, informed, and in-depth analysis and commentary on the Aurea Catena Homeri.

It was also highly recommended by one of the most informed past members of this Forum, Green Lion.

I think that it’s vital that this treatise be translated. As far as I’m aware, there does not exist a translation of this German Treatise.

The manuscript in this thread is quite substantial, and consists of 192 pages! That is quite a large work.

Since 2017, when this thread was active. Prices have soared, and today a professional translator would charge $30-75 USD per page. Assuming an average of $50 per page, that would mean a total of $9,600. That’s a lot and would mean that 192 members would have to chip in $50 USD each to cover the cost. Obviously, that is not going to happen.

So, what options do we have?

We could go to an experienced amateur, whose product might not be perfect, but would be good enough to get a fairly good and acceptable understanding of the work. An amateur might only charge half of this price, so approximately $5,000. That would still mean that 100 members would have to contribute $50 each.

That still does not seem doable. I have never seen more than about 10-20 members willing to chip in for a work, even if that work might be extremely important such as this.

So, what do we do?

I would be willing to do this translation work, but even at $5,000 that is too much for that to happen in this group. Because of my charitable nature, even though this involves substantial work, I could do the work for the ridiculously low price of $1,000, a fifth of the already greatly reduced price. At $1,000. It would only require 20 members here, each providing $50. That might just be doable.

But hey, I mentioned that I was a charitable person, and I mean it. I would be insane to do all this work for $1,000 and even more insane to do it for a tenth of that price - $100 or so. Who in their right mind would do all that work for perhaps $2.00 an hour?

At $100 we only need about seven people contributing not $50 each but $15 each.

This is a total steal of a deal. Please let me know if you are interested, and if I don’t come to my senses before then, I can actually make this happen, although at this price, the resultant work won’t be perfect. Many of you know that I’m capable of this and can provide a reasonably good result.

I might need a little help, though. If you are experienced with German Fractur and are an experienced and careful proof reader, let me know, and the translation can be yours for free. I'll take care of the initial transcription, and it will probably be at least 99% accurate, but proof-reading can increase the accuracy. The more proof-readers, the better.

Let me know soon before I change my mind.

 

[Philosophical]

I know that many here would fully agree that Aurea Catena Homeri (Golden Chain of Homer) is considered to be one of the most descriptive and least ambiguous manuscripts of alchemical lore. It is more focused on outright experimentation, and covers the processes of fixation, putrefaction, and generation in far more depth than most works of the era. It moves away from the more destructive alchemical approaches involving such ingredients as Antimony, and seeks a far more natural process of achieving the Philosophers Stone that involves man’s urine. The Golden Chain is a similar to the approach used by other Alchemists such as Goosen Van Vreeswyk.

The Treatise mentioned in this thread seems even more valuable, as it actually provides a very serious, informed, and in-depth analysis and commentary on the Aurea Catena Homeri.

It was also highly recommended by one of the most informed past members of this Forum, Green Lion.

I think that it’s vital that this treatise be translated. As far as I’m aware, there does not exist a translation of this German Treatise.

The manuscript in this thread is quite substantial, and consists of 192 pages! That is quite a large work.

Since 2017, when this thread was active. Prices have soared, and today a professional translator would charge $30-75 USD per page. Assuming an average of $50 per page, that would mean a total of $9,600. That’s a lot and would mean that 192 members would have to hip in $50 USD each to cover the cost. Obviously, that is not going to happen.

So, what options do we have?

We could go to an experienced amateur, whose product might not be perfect, but would be good enough to get a fairly good and acceptable understanding of the work. An amateur might only charge half of this price, so approximately $5,000. That would still mean that 100 members would have to contribute $50 each.

That still does not seem doable. I have never seen more than about 10-20 members willing to chip in for a work, even if that work might be extremely important such as this.

So, what do we do?

I would be willing to do this translation work, but even at $5,000 that is too much for that to happen in this group. Because of my charitable nature, even though this involves substantial work, I could do the work for the ridiculously low price of $1,000, a fifth of the already greatly reduced price. At $1,000. It would only require 20 members here, each providing $50. That might just be doable.

But hey, I mentioned that I was a charitable person, and I mean it. I would be insane to do all this work for $1,000 and even more insane to do it for a tenth of that price - $100 or so. Who in their right mind would do all that work for perhaps $2.00 an hour?

At $100 we only need about seven people contributing not $50 each but $15 each.

This is a total steal of a deal. Please let me know if you are interested, and if I don’t come to my senses before then, I can actually make this happen, although at this price, the resultant work won’t be perfect. Many of you know that I’m capable of this and can provide a reasonably good result.

I might need a little help, though. If you are experienced with German Fractur and are an experienced and careful proof reader, let me know, and the translation can be yours for free. I take care of the initial transcription, and it will probably be at least 99% accurate, but proof-reading can increase the accuracy. The more proof-readers, the better.

Let me know soon beffore I change my mind.

You know I'm always down for one of your translations @Illen. You are so generous with your translations and I wish there were 200 of us who value what you do to chip in the same amount to make it more worth your while.

1/20

 

[Andro]

I can chip in as well.

Are you certain it's discussing a urine process?

Do you have a good quality scan of the Old German original?

 

[Illen A. Cluf]

I can chip in as well.

Are you certain it's discussing a urine process?

Do you have a good quality scan of the Old German original?

Until the translation is complete, I'm not sure what it contains. But I believe the Aurea Catena Homeri involves the urine process, and this is supposed to be a commentary/analysis of that.

Yes, the scan is quite adequate. I can do a few pages and present it here for observation.

 

[Lux Natura]

AI aided OCR tools are excellent. In the past, after OCR (with the output matches the input line for line), I just place the two documents beside each other and check each letter / word as I go, and make corrections if there are any issues.

All the translations I do via AI, some models are better than others. I've learned to have it also highlight or signal sections in which the LLM is not 100% confident in the translation, or where other possibilities may exist, so I can manually research and recheck this, often with searches online for similar phrases. That's how, for example, I mistranslated the amount of material needed to work with to 1000kg (a ton), where it should have read "a barrel / bucket" (this was for ICH's other work). Thankfully Andro knew right away, and with further reading I understood the reasons for the mistranslation.

I mention all this because a "pretty good" machine aided translation can be had in a few hours or a day with the proper setup, also depends on the length of the work. Maybe not as good as a dedicated expert that can understand the language and its nuances, but good enough to get most things correct at least.

Some preliminary setup to create global rules and things like detailing the unicode sets used for symbols and the like is important, as well as instructing the LLM to ignore catchwords at the end of pages, keep the Latin untranslated, and so on.

 

[Appleheads]

I will chip in also

 

[Lux Natura]

I should have mentioned, I will chip in too - and if you don't get the 7 people you need to cover $15 each, I will cover the rest.

 

[ghetto alchemist]

AI aided OCR tools are excellent.

Can you get one of these AI OCR tools that you speak of, to run through the images on THIS page.
Then we can take a look and see if its any good at olde german.
Looking beyond the scope of this project, I think having a pipeline for machine transcribing and translating old german alchemy texts would be worth having on its own.
Then you can use the AI transcribed/translated result as the baseline to get even cheaper translation fees.
Eg: it's gotta be easier for a translator to correct small errors in a machine translation than start from scratch on his/her own effort.

 

[Mahavira]

I'm in.

 

[Lux Natura]

Can you get one of these AI OCR tools that you speak of, to run through the images on THIS page.

AI OCR is basically trivially easy at this point. It's the translations that are sometimes lacking, but in other cases they can be good enough.

I randomly chose a page, to show proof of concept - the images are 1: original page, 2: machine OCR, 3: AI word-for-word translation, 4: AI human readable translation.

It took longer to take the screenshots and put them in an image than actually doing the "AI work". Bear in mind, this is without doing any checking to see if the OCR is accurate or translation is accurate. You are free to check to see (it's page 145). I think you will find it to be pretty close, if not totally accurate as far as OCR. The translation I cannot say how good or bad it is.

 

[Lux Natura]

I just wanted to add a small thing on the topic of whether or not AI is good enough to do OCR / translation, and I don't mean to distract from the main topic of this thread, just add a small update in response to @ghetto alchemist question.

As a general proof of concept, you can check out https://sourcelibrary.org/ as it contains a large collection of purely machine translated texts, from a wide variety of topics, including alchemy. When you go to read a book, it will show you the original scan, the OCR, and the translation, and if you choose the "trace" option, you can highlight a section in one, and see the corresponding section in another. There are plenty of texts there in German, using the Fraktur typeface. There are about 1300 alchemical books there translated, and many are the only available translations that exist.

 

[Andro]

I randomly chose a page, to show proof of concept - the images are 1: original page, 2: machine OCR, 3: AI word-for-word translation, 4: AI human readable translation.

OCR/character recognition is OK. The "human-friendly" rendering, while containing few (but enough) inaccuracies to somewhat mislead, should be enough for a Fraktur-to-English proof reader with a more-than-superficial background in Alchemy.

From this one page only, it appears to be a Circulatum type of process. The Spirit (volatile salts) are extracted by distillation, so is the oil ("Soul of Mars"), and then the fixed salt is extracted by lixiviation and subsequently united with its own Spirit over 40 days.

This generalization is only based on the one page posted above, I haven't checked any other pages. Reading Fraktur makes my head spin.

 

[Illen A. Cluf]

AI OCR is basically trivially easy at this point. It's the translations that are sometimes lacking, but in other cases they can be good enough.

I randomly chose a page, to show proof of concept - the images are 1: original page, 2: machine OCR, 3: AI word-for-word translation, 4: AI human readable translation.

It took longer to take the screenshots and put them in an image than actually doing the "AI work". Bear in mind, this is without doing any checking to see if the OCR is accurate or translation is accurate. You are free to check to see (it's page 145). I think you will find it to be pretty close, if not totally accurate as far as OCR. The translation I cannot say how good or bad it is.

View attachment 5061

What tools did you use to do this, @Lux Natura ?

 

[Lux Natura]

From this one page only, it appears to be a Circulatum type of process. The Spirit (volatile salts) are extracted by distillation, so is the oil ("Soul of Mars"), and then the fixed salt is extracted by lixiviation and subsequently united with its own Spirit over 40 days.

I "spot tested" a few other parts yesterday. Of course I cannot tell how good the translation is because I do not speak or read German. An earlier section deals with using natural sulfides, pounding them into a powder, dry distilling to get an oil (which is first the white like smoke then red like blood). The white and red will separate or form layers when it cools, and they are kept separate. The red is called the acidic metallic spirit from Venus / Mars. The white is called the central spirit (spiritus centralis).

A preparation of stibnite ore and iron is used to create the regulus martis, onto which is poured the white spirit, and over time, under a gentle fire you get a sublimation of blue green crystalline vitriol in the neck of the retort. This sublimate is identified with the mercury of the philosophers.

It is then dissolved in rain water and sealed for 40 days in warmth to undergo putrefaction, turning into a pitch black stinking water, where it is called the materia prima or chaos philosophorum. Once the time has passed, it is distilled, to get a clear liquor from it, which is called the mercurius vegetabilis.

Remaining in the bottom from the putrefaction, the black earth is calcined until snow white, and from there a salt is extracted using the mercurial solvent from above, which is coagulated into the crystal grains. Later on, these crystal grains and the mercury from above are united in a certain ratio (1:3) in the sealed egg, which undergoes the blackness, peacock colors, white, then finally red.

The multiplication of this occurs using the initial red spirit.

That is what I remember of one of yesterdays spot test "translations". That was done without any further checking for errors or anything of the like, so it may be entirely incorrect, just giving an example of some other part. I believe it was the first or second "chapter" of the book.

 

[Lux Natura]

What tools did you use to do this, @Lux Natura ?

That was entirely with google Gemini set on fast, but there are a bunch of global rules I have set for translation workflows, otherwise it like to hallucinate too much, or do its own thing and fail to follow instructions. It's best to feed it 1 page at a time. You can get much better workflows if you can use agents (AI's that oversee other AI's, hyperspecific roles for certain AI's, etc, exhaustive re-checking, etc).

The workflow I've used in the past splits the pdf into individual images, and sequentially numbers them. If the images are bad, there are python tools to straighten, crop, color correct, etc. The output of the OCR captures what image it was, as well as the page number, so that rechecking later is easier. One can even use a deeper level of the LLM to do the rechecking, letter by letter if one wants, or even a totally other model of AI.

The translation here is with Gemini flash, but the other "reasoning" models may be more accurate. I have used a bunch of others in the past, such as deepseek. With Gemini it's best to just work 1 page at a time though.

 

[Illen A. Cluf]

That was entirely with google Gemini set on fast, but there are a bunch of global rules I have set for translation workflows, otherwise it like to hallucinate too much, or do its own thing and fail to follow instructions. It's best to feed it 1 page at a time. You can get much better workflows if you can use agents (AI's that oversee other AI's, hyperspecific roles for certain AI's, etc, exhaustive re-checking, etc).

The workflow I've used in the past splits the pdf into individual images, and sequentially numbers them. If the images are bad, there are python tools to straighten, crop, color correct, etc. The output of the OCR captures what image it was, as well as the page number, so that rechecking later is easier. One can even use a deeper level of the LLM to do the rechecking, letter by letter if one wants, or even a totally other model of AI.

The translation here is with Gemini flash, but the other "reasoning" models may be more accurate. I have used a bunch of others in the past, such as deepseek. With Gemini it's best to just work 1 page at a time though.

That's what I have been using for the first few pages - one page at a time. I think it does a reasonable enough job, but if we can find some members who are willing to check the transcrition, that would be better.

 

[Lux Natura]

Once you have an OCR you can cross reference the translation using different LLM models. Unless the text is purposefully using unique or convoluted wordplay, codenames, or something of the like, the reliability of it translating simple words and phrases is pretty good. There is a lot you can do to help automate tasks and streamline workflow - I usually work with 2-4 instances of the "AI" being open at once, each one being tasked with doing specific things. Then mostly the role is to copy and paste the relevant things you want it to process, and then save the outputs into a document that you ultimately end up rechecking / editing. One could actually even automate many of the copy/paste tasks too with some simple code.

The workflow can be pretty quick once this is set up - we talked about this in private messages in the past. I've done some 200+ page texts in the course of a few days, maybe not good enough for "professional high grade translation", but certainly good enough for a basic personal reading.

 

[Illen A. Cluf]

Once you have an OCR you can cross reference the translation using different LLM models. Unless the text is purposefully using unique or convoluted wordplay, codenames, or something of the like, the reliability of it translating simple words and phrases is pretty good. There is a lot you can do to help automate tasks and streamline workflow - I usually work with 2-4 instances of the "AI" being open at once, each one being tasked with doing specific things. Then mostly the role is to copy and paste the relevant things you want it to process, and then save the outputs into a document that you ultimately end up rechecking / editing. One could actually even automate many of the copy/paste tasks too with some simple code.

The workflow can be pretty quick once this is set up - we talked about this in private messages in the past. I've done some 200+ page texts in the course of a few days, maybe not good enough for "professional high grade translation", but certainly good enough for a basic personal reading.

I plan to do this in a lot less time and effort than I would normally do for a book. To do a professional job would require a 100 hours or more, even with today's aids. I plan on spending half that amount. The final prouct should be sufficient enough. I can always go back and do a better job on any issues that come up.

 

[Lux Natura]

I plan to do this in a lot less time and effort than I would normally do for a book. To do a professional job would require a 100 hours or more, even with today's aids. I plan on spending half that amount. The final prouct should be suficient enough. I can always go back and do a better job on any issues that come up.

Yeah, exactly. With the tools available, others can always choose to go back and scrutinize certain sections in more details if they want. If its at all like most texts, much of it will be straight forwards, minus a few small things that may end up being code words, or code processes, and those are always the lynch pin behind it all. It's also useful to more "quickly" translate things to get an idea or feel for the text to find out if in fact it deals with the kinds of things one is interested in. Just because something is old and was written a long time ago does not mean that the authenticity of it being an 'aged text' also somehow magically confers an authenticity to the processes it deals with.

I have found that some alchemical processes I have discovered have exactly zero reference in texts, unless I am to take a very generous and wide berth in interpretation, but at that point, you can pretty much interpret anything with a wide enough viewpoint so it can mean whatever you want it to mean.

 

[Illen A. Cluf]

Yeah, exactly. With the tools available, others can always choose to go back and scrutinize certain sections in more details if they want. If its at all like most texts, much of it will be straight forwards, minus a few small things that may end up being code words, or code processes, and those are always the lynch pin behind it all. It's also useful to more "quickly" translate things to get an idea or feel for the text to find out if in fact it deals with the kinds of things one is interested in. Just because something is old and was written a long time ago does not mean that the authenticity of it being an 'aged text' also somehow magically confers an authenticity to the processes it deals with.

I have found that some alchemical processes I have discovered have exactly zero reference in texts, unless I am to take a very generous and wide berth in interpretation, but at that point, you can pretty much interpret anything with a wide enough viewpoint so it can mean whatever you want it to mean.

I did a few pages and was impressed by the results. I also found that even if there is a slight mis-transcription, when the translation is done, it picks up the error -- in context -- and adjusts for it, especially when specially advised. So for this exercise, I think an AI transcription and an AI translation should be more than adequate and give us a result in a relatively short time. The AI translators seem be be improving by the month. Once we read the final result, we can likely easily spot most remaining errors.

 

[Lux Natura]

I did a few pages and was impressed by the results. I also found that even if there is a slight mis-transcription, when the translation is done, it picks up the error -- in context -- and adjusts for it, especially when specially advised. So for this exercise, I think an AI transcription and an AI translation should be more than adequate and give us a result in a relatively short time. The AI translators seem be be improving by the month. Once we read the final result, we can likely easily spot most remaining errors.

My man! You are coming around to my style of thinking! This is what I was saying in reference to our discussion previously on Goossen van Vreeswijck's works. There will always be a place for a proper in-depth translation done with the utmost attention and care, but the machine translations are not bad, often "good enough" for most purposes, and improving by leaps and bounds as time passes.

 

[ArgentSeeker]

That's what I have been using for the first few pages - one page at a time. I think it does a reasonable enough job, but if we can find some members who are willing to check the transcrition, that would be better.

Translating and Indexing Manuscripts

Hello. I've made a guide on how you can do completely private indexing, translation, and adjustment of any scanned book or manuscript. It uses all free and open source models and programs to achieve a number of tasks. Here's two guides, and example of in-progress work. Obviously, this doesn't...

alchemyforums.com

I provided a guide on how to do this here. I will help and work with you, if you want. I will also proof-read. Let me know if you want help or would like me to help you make a pipeline. I can make a whole setup for you so that everything is processed privately on your computer and nothing goes out as well, you just need a good gpu and a good cpu, or you'll have to wait longer.

My man! You are coming around to my style of thinking! This is what I was saying in reference to our discussion previously on Goossen van Vreeswijck's works. There will always be a place for a proper in-depth translation done with the utmost attention and care, but the machine translations are not bad, often "good enough" for most purposes, and improving by leaps and bounds as time passes.

I think potentially preferred procedure would be to do a rough-draft, or general translation, using AI, then going back and confirming line by line. That makes the process and transcription much faster for basic, common sentences, and allows the translator to focus on the more complex or obscure phrases. This way it's still human-confirmed, but the machine translation does most of the heavy lifting.