Morgan M.873 showed the material world entering medicine.

Masson 116 showed a medical image library surviving after its extensive explanatory layer failed to arrive.

The Carrara Herbal joined medicinal substances to pharmacological meaning in the language and political environment of Padua.

Sloane MS 4016 reorganised the visual corpus into a deliberate image-and-name interface.

The next manuscript takes us upstream.

Bibliothèque nationale de France, Latin 6823, associated with Manfredus de Monte Imperiale, preserves a world in which many of those functions remain connected.

It contains an illustrated work on medicinal simples, but the present codex does not stop there. It also contains material concerning animals, urine, compound medicines, doses, substitutions, additional recipes and medical synonyms.

This changes its position in the Atlas.

It is not merely an earlier herbal from which later pictures descended.

It is a surviving approximation of an entire medical knowledge stack.

Recognise the substance
→ understand its properties
→ extend the inventory beyond plants
→ read signs from the body
→ formulate compound remedies
→ determine quantities
→ replace unavailable ingredients
→ resolve conflicting names

The later manuscripts preserve and redeploy individual parts of this structure.

Latin 6823 allows us to see what the structure looked like while many of its parts were still assembled inside one codex.

The first necessary distinction

The name Manfredus Herbal can refer narrowly to the Tractatus de herbis et plantis attributed to Manfredus.

But Latin 6823 is the larger physical codex in which that work now survives.

The difference is important.

The work

Manfredus’s Tractatus de herbis et plantis occupies folios 3r–170v. Its opening says that Manfredus wished to learn the virtues and properties of simple medicines and that he recorded and demonstrated them through figures.

The codex

The present manuscript contains 249 parchment folios measuring approximately 345 by 250 millimetres. The BnF catalogue dates it broadly to 1301–1350, records Italian Gothic script and notes painted illustrations on nearly every folio from 1 to 179. The present catalogue gives Pisa as the place of production. Other art-historical and scientific scholarship has instead associated its prefatory illumination with Naples or southern Italy around 1330–1340. The safest coordinate is therefore Italy, first half of the fourteenth century, with the precise production centre still disputed.

The larger codex contains:

• Manfredus’s illustrated work on herbs and plants;
• a work on animals, birds and fish;
• short medical experiments;
• a work on urine associated with Constantine the African;
• the Antidotarium Nicolai;
• a section on doses;
• a Quid pro quo on substitutions;
• additional medical preparations;
• an extensive collection of medical synonyms associated with Simon of Genoa and Mundinus.

That does not automatically prove that all these texts were conceived by one person, produced during one campaign or used as one carefully planned manual.

The codex may have grown through compilation, addition and rebinding.

But whether intentionally designed from the beginning or historically assembled, its present architecture creates something close to an operational medical library.

The book begins with a living interface

Before the alphabetical medical entries begin, the manuscript presents a remarkable scene.

A seated bearded figure, generally understood as Manfredus, faces men who hold up plants for examination. The image can be read as teaching, identification or consultation: physical specimens are brought before someone capable of connecting visible matter to learned knowledge.

The following pages depict pairs of medical authorities in conversation. They include figures identified or interpreted as Hippocrates, Galen, Ḥunayn ibn Isḥāq—known in Latin as Johannitius—Mesue, Bartholomew of Salerno, Averroes and Porphyry. Their words cross the space between them, creating conversations between people separated by centuries, languages and geography.

The opening sequence gives us two forms of knowledge production.

Living specimen
→ teacher
→ student

and:

Earlier authority
→ translation
→ commentary
→ later compiler

Manfredus stands between them.

He is not presented merely as someone who repeats an old text.

He receives plants from the material world while also receiving medicine from an inherited intellectual world.

The manuscript’s operating principle appears before its first main entry:

Medical knowledge is created by keeping the object, the name, the image, the authority and the trained interpreter connected.

A classroom wider than history

The pictured authorities could never have occupied one physical room.

Some belonged to classical Greece.

Some transmitted or transformed Greek medicine through Arabic.

Some belonged to the Salernitan medical tradition.

Some were philosophers and commentators whose works entered medieval medical education.

Yet the manuscript installs them together.

This is not historical confusion.

It is an intellectual architecture.

Hippocrates
        ↓
Galen
        ↓
Greek and Syriac transmission
        ↓
Arabic medicine and commentary
        ↓
Latin translation
        ↓
Salernitan teaching
        ↓
Manfredus
        ↓
student and reader

The exact path differs for each text and authority, but the image expresses the larger structure: medicine is cumulative, translated and collaborative.

The manuscript page becomes a room in which the entire medical inheritance can speak at once.

Sarah Kyle interprets these prefatory conversations as occurring in the mind of Manfredus and of the attentive reader. They present the compiler as a participant in a continuing lineage rather than as the isolated originator of the knowledge.

This is one of the first major connections the manuscript adds to our growing world.

The medical book was not simply a database of substances.

It was also an authority-routing system.

Text and image remain joined

In Masson 116, most of the images survive while the extensive explanatory text that its page design appears prepared to receive is absent.

In Sloane 4016, the visual layer becomes an intentional full-page interface.

Latin 6823 preserves an earlier configuration.

The plant is painted.

The name is supplied.

The surrounding writing explains properties, qualities or medical associations.

The reader does not have to choose between a verbal pharmacology and a visual catalogue. The two systems occupy the same working surface.

Text asks:
What do authorities say about this substance?
Image asks:
What visible object is this entry attempting to represent?

The conjunction matters because each channel repairs a weakness in the other.

A name can travel far from the object it originally identified.

An image can become ambiguous without a label.

A written description can be interpreted differently by readers in different regions.

A copied picture can drift across generations.

When text and image remain together, they create a form of mutual constraint.

[
\text{Entry stability}

\text{Name}
+
\text{Image}
+
\text{Description}
+
\text{Position}
]

No component guarantees correctness.

Together they reduce the range of possible error.

The manuscript is not organised as modern botany

The central question of the book is not:

What species belongs in which biological family?

Its working question is closer to:

What substance is recognised by medicine, what is it called, what properties are attributed to it, and how can it be represented?

Plants dominate much of the illustrated sequence, but the larger codex extends into animals, minerals, manufactured materials and medical preparations.

The BnF’s manuscript description indexes scenes showing the extraction of alum, antimony, orpiment, Armenian bole and sulphur; the collection of aloe wood and quicksilver; the production of glass objects; and numerous plants and animal materials.

The manuscript therefore does not divide the medical world according to modern scientific departments.

Its structure is organised around therapeutic availability.

Plant
Animal
Mineral
Extracted material
Manufactured substance
Compound preparation

All can enter one medical system because all can become relevant to maintaining, diagnosing or treating the body.

The book contains the upstream world

Morgan M.873 made the supply universe especially visible through its images of collection, extraction and manufacture.

Latin 6823 contains many of the same kinds of upstream scenes.

The manuscript shows that a medical entry begins before the apothecary receives a drug.

A substance may have to be:

• cultivated;
• gathered;
• hunted;
• extracted;
• processed;
• transported;
• manufactured.

The picture of a completed plant is therefore only one coordinate.

The full material route is:

[
\text{environment}
\rightarrow
\text{source object}
\rightarrow
\text{harvest or extraction}
\rightarrow
\text{commercial substance}
\rightarrow
\text{medical simple}
]

The visual system occasionally exposes these transformations.

More often, the finished entry compresses them.

A reader sees the medicinal object after the production cloud has disappeared into the page.

Then the codex crosses from world to body

The most important discovery begins after the illustrated materia medica.

The present manuscript does not end when its sequence of plants and other natural substances ends.

It turns toward the human body.

A short work De urinis, associated with Constantine the African, appears at folios 181r–182v.

Urine was treated within medieval learned medicine as a visible bodily signal. Its colour, clarity, sediment and other qualities could be interpreted within an inherited diagnostic framework.

We do not need to accept the medical validity of every historical interpretation to see the system architecture.

The direction of movement has changed.

First half of the system:
World → substance
Next layer:
Body → visible signal

The medical practitioner now stands between two information streams:

External input:
What medicinal materials are available?
Internal signal:
What appears to be happening in the body?

A remedy can be chosen only after these two sides are brought into relation.

The manuscript therefore begins to bridge materia medica and diagnosis.

The Antidotarium introduces compound medicine

At folio 185, the codex begins the Antidotarium Nicolai, a highly influential Salernitan collection of compound remedies. The section continues to folio 200v and is introduced by an illuminated initial depicting Nicolas teaching.

This marks another major transition.

A simple medicine is treated as an individual substance.

A compound medicine is assembled from several materials.

The information grammar changes.

Simple entry

Substance X
has properties A, B and C.

Compound formula

Take X.
Add Y.
Prepare Z.
Use specified quantities.
Transform them through a sequence.
Produce remedy R.

The medical book is no longer only describing what exists.

It is instructing the reader how to produce something that does not yet exist.

That is the passage from inventory to manufacture.

[
\text{Materia medica}
\rightarrow
\text{selection}
\rightarrow
\text{combination}
\rightarrow
\text{compound remedy}
]

The codex now contains both the component library and an assembly layer.

Dose changes knowledge into controlled action

After the Antidotarium, Latin 6823 contains a section explicitly concerned with doses.

Its opening explains that after discussing the preparation of compounds, plasters and ointments, attention must turn to the quantities of spices and herbs.

This is a critical boundary.

Knowing the correct substance is not enough.

Knowing the correct preparation is not enough.

The amount matters.

Correct material
+
correct method
+
wrong quantity
=
incorrect execution

Dose converts qualitative knowledge into controlled action.

It introduces measurement into the runtime.

The practical chain becomes:

Identify
→ select
→ prepare
→ measure
→ administer

A picture book alone cannot reliably perform this function.

A pharmacological description alone may not perform it either.

Dose belongs to another module.

Latin 6823 places that module beside the others.

Quid pro quo repairs a broken supply chain

Immediately after the section on doses comes a short work known as Quid pro quo.

Its opening addresses the problem of placing one substance in place of another when the required material cannot be found.

This is not a minor appendix.

It responds to one of the most serious operational problems in premodern medicine:

What happens when the prescribed ingredient is unavailable?

A remedy may fail before preparation begins because of:

• season;
• distance;
• interrupted trade;
• scarcity;
• cost;
• spoilage;
• political conflict;
• uncertain naming;
• local absence.

The substitution list introduces recoverability.

Required substance unavailable
→ locate permitted substitute
→ preserve as much of the intended function as possible
→ continue preparation

This is a resilience module.

Without it, a medical formula is brittle.

It works only when every specified input is present.

With a Quid pro quo, the system can continue operating under imperfect supply conditions.

The manuscript therefore contains not only a nominal recipe system, but a primitive repair architecture.

Additional recipes prevent the canon from becoming closed

The manuscript next includes a collection of preparations explicitly described as medicines and compounds not found in the standard Antidotarium Nicolai.

The section occupies folios 203v–214r and includes preparations such as medicinal vinegar and other formulations.

This creates another important distinction.

Canonical formulary
≠
complete universe of practice

A recognised reference work can provide a stable core.

Practice continues outside it.

Additional recipes may arise from:

• other textual authorities;
• local traditions;
• professional experience;
• household medicine;
• regional ingredients;
• later correction or expansion.

The manuscript therefore preserves both standardisation and extension.

This is how a practical knowledge system avoids two opposite failures.

Without a stable canon

Every practitioner may use incompatible formulas.

Without an extension mechanism

The system cannot absorb new experience or local variation.

Latin 6823 carries both:

Antidotarium
=
protected common core
Additional confectiones
=
open extension layer

The synonym index resolves the final translation problem

The codex concludes with a substantial collection of medical synonyms associated with Mundinus of Friuli and ultimately with the lexicographical tradition of Simon of Genoa.

The section occupies folios 215r–249r.

Its position is conceptually powerful.

The manuscript has already supplied:

• substances;
• images;
• properties;
• animal materials;
• bodily signals;
• compounds;
• doses;
• substitutions;
• additional recipes.

But the whole machine can still fail if the operator cannot resolve a name.

A medicinal material may carry:

Greek-derived name
Arabic-derived name
Latin scholarly name
regional name
merchant’s name
shop synonym
scribal variant

Simon of Genoa’s medical lexicographical tradition was specifically concerned with terminology moving among Latin, Greek and Arabic medical worlds.

The synonym list acts as a translation layer.

[
\text{Word in source}
\rightarrow
\text{recognised equivalent}
\rightarrow
\text{physical substance}
]

Without this layer, a perfectly copied recipe could still be operationally useless.

The reader may understand the instruction but fail to know which object the instruction names.

The medical runtime therefore ends where it began:

Beginning:
A plant is presented for identification.
End:
A word is presented for resolution.

The whole codex is enclosed by the problem of matching representation to reality.

The complete stack inside Latin 6823

The present manuscript can now be mapped as a sequence of operating layers.

This is not yet a complete hospital, pharmacy or clinical consultation.

But it is much closer to an executable medical system than any one of its later visual descendants.

The codex contains:

Objects
Names
Properties
Diagnosis signals
Recipes
Quantities
Substitutions
Extensions
Translation

What remains outside includes:

• the individual patient;
• physical examination;
• the practitioner’s final judgement;
• material-quality testing;
• preparation skill;
• timing;
• contraindications;
• observation of outcome;
• correction after failure.

The book supplies a large part of the architecture.

The trained human still supplies the runtime.

A book for one profession—or a meeting surface?

It would be tempting to declare Latin 6823 an apothecary’s manual.

Its contents clearly overlap with pharmaceutical work.

Manfredus’s opening associates him with the desire to understand the virtues and properties of simples within the ars speciaria, a term connected to the world of medicinal substances and spices.

But one label would be too narrow.

The codex could speak to several roles.

The physician

Uses learned medicine, bodily signs and pharmacological properties to make therapeutic decisions.

The apothecary

Connects textual names to physical stock, judges availability, measures ingredients and prepares compounds.

The student

Learns the recognised substances, authorities, recipes and vocabulary of medicine.

The teacher

Demonstrates how visible specimens connect to inherited texts.

The compiler

Collects dispersed authorities and places them in an ordered system.

The illustrator

Converts substances, extraction scenes and medical concepts into reusable visual forms.

The patron or institutional owner

Funds, possesses and preserves the expensive object.

The later copyist

Extracts images, captions or sequences for installation in another manuscript.

The codex may have supported several of these relationships without belonging exclusively to one.

Owner
≠
reader
≠
medical operator
≠
copyist

That separation must remain intact.

The production cloud

Latin 6823 required an unusually wide convergence of capabilities.

Textual inheritance

The codex brings together medical traditions associated with Salerno, Constantine the African, Nicolas of Salerno, Simon of Genoa and a much older Greek–Arabic–Latin chain of authorities.

Material knowledge

The compiler needed access to recognised accounts of plants, animals, minerals and prepared substances.

Visual labour

Hundreds of illustrations had to be planned, drawn, coloured and coordinated with writing.

Scribal labour

Different texts, rubrics, initials, entries and supplementary materials had to be copied and organised.

Pharmaceutical understanding

The system had to distinguish simples, compounds, quantities, substitutes and synonyms.

Institutional surplus

Someone had to finance the parchment, pigments, time and skilled labour.

The Production Energy can be represented as:

[
P_E =
T + V + M + L + A + C + S
]

Where:

• (T) = textual sources;
• (V) = visual models and observation;
• (M) = medical and material expertise;
• (L) = labour;
• (A) = access to authorities and libraries;
• (C) = coordination;
• (S) = economic surplus.

No component is individually impossible.

The achievement lies in their convergence.

The book migrates north

The manuscript’s earliest geographic coordinate remains contested.

The present BnF catalogue records Pisa, while significant scholarship places its artistic production in Naples or southern Italy around the 1330s.

Its later visual descendants, however, clearly appear in northern Italy.

This movement matters more than choosing one unsupported early route.

A southern or central Italian medical compilation could become a source engine within the northern manuscript world.

Possible southern production
→ northward movement
→ access by new workshops
→ extraction of images
→ new medical books

Later inventory scholarship identifies a copy of Manfredus’s work in the 1426 library catalogue of the Visconti and Sforza dukes of Milan, and has associated that entry with the present Latin 6823. The same inventory contained another Manfredus copy that is now lost. The identification is an important scholarly reconstruction rather than a complete custody trail for every intervening year.

By that stage, the manuscript was no longer merely an author’s or teacher’s working compilation.

It had entered a princely library capable of supporting further manuscript production.

The Masson extraction

Masson 116 now becomes easier to understand.

Biblissima records that most of Masson’s plant iconography was copied from the Tractatus de herbis of Manfredus in Latin 6823. Its animals and figurative scenes often derive from the animal-medicine material that follows Manfredus’s herbal in the same codex.

This means Masson’s makers did not merely copy “plant pictures.”

They extracted modules from different parts of a larger medical assembly.

Latin 6823 contains:
text + plant image + animal medicine + recipes + synonyms
Masson extracts mainly:
plant image + animal scene + label

The downstream manuscript therefore preserves only selected layers of the upstream system.

That is why Masson’s silence is so revealing.

The missing explanatory world is not hypothetical.

A version of that world survives in its source environment.

Masson is what happens when the visual layer is detached and recommissioned.

Morgan as the input projection

Morgan M.873 can also be repositioned.

Its hundreds of plants, animals, minerals and scenes of extraction or manufacture resemble the material-input face of the system.

Morgan is especially strong at:

What enters medicine?
Where might it originate?
What does the material or source object look like?

Latin 6823 shows how that input layer can sit beside:

• longer explanations;
• compound formulas;
• quantities;
• substitutions;
• synonym resolution.

Whether every Morgan image passed directly through the surviving Paris codex requires image-by-image transmission research.

Functionally, however, Morgan can now be recognised as a projection of one region of the larger stack.

It preserves the supply and recognition layer while much of the verbal execution layer is absent.

Sloane as the retrieval projection

Sloane 4016 takes another region.

It enlarges the image-and-name relationship and turns it into the dominant page architecture.

Latin 6823:
image embedded in an extensive medical compilation
Sloane 4016:
image and caption become the primary lookup surface

The later manuscript does not reproduce the entire upstream runtime.

It selects the visual retrieval system.

This is not necessarily degeneration.

It may be optimisation for a different job.

A system designed to support recognition, reference, display or further copying does not need to reproduce every recipe and dose beside every image.

Sloane demonstrates that one medical stack can produce specialised downstream instruments.

Carrara as local reinstallation

The Carrara Herbal performs another transformation.

It takes pharmacological knowledge and installs it in a Paduan vernacular court manuscript.

It adds:

• local language;
• court patronage;
• political meaning;
• a different pharmacological source structure;
• a new visual programme.

Manfredus and Carrara are therefore not interchangeable texts.

Yet Latin 6823 helps explain the environment in which a later Paduan ruler could see value in uniting inherited medicine, images and local language.

The earlier codex demonstrates that:

authority
+
image
+
materia medica
+
teaching

could already function as one system.

The Carrara project recompiled that possibility for another city, patron and political purpose.

The hidden law of the manuscript family

We can now identify a larger pattern.

The later books are not simply inferior copies losing information at random.

They may be specialised decompressions of different components from an earlier, denser stack.

Latin 6823
=
large integrated medical knowledge environment

From it or from its wider tradition, later manuscripts emphasise different functions:

Morgan M.873
=
material-input and procurement interface
Masson 116
=
visual component library
Sloane 4016
=
image–name retrieval interface
Carrara Herbal
=
vernacular pharmacological and political installation

Each appears incomplete when judged against the total system.

Each may be coherent when judged against its selected job.

This produces an Atlas law:

Apparent manuscript incompleteness may result from functional specialisation rather than simple knowledge loss.

The correct question is not only:

What is missing from this book?

It is also:

Which layer of the larger system was this book built to carry?

The remaining void changes shape

Before Latin 6823 was placed into the map, the missing region seemed to lie broadly between medicinal image and executed remedy.

The codex fills much of that space.

It gives us:

• properties;
• body-side diagnostic material;
• compound formulas;
• quantities;
• substitutions;
• recipes beyond the standard text;
• terminology resolution.

The remaining void is now smaller and more human.

Book-side knowledge
→ trained judgement
→ material execution
→ patient response

What is still missing is not primarily another catalogue of plants.

It is the daily operating layer that connected several books and materials:

• deciding which text applied to the present patient;
• judging the quality of the available drug;
• adjusting for age, strength and circumstance;
• preparing a substance correctly by hand;
• observing whether the remedy helped or harmed;
• changing course when treatment failed.

Those functions may have been distributed among physician, apothecary, assistant, household and patient.

They may never have been fully captured in one luxury manuscript because they belonged to living practice.

Latin 6823 does not eliminate the apparent void.

It relocates it.

The gap now lies less between book and book and more between book and trained human action.

Production Energy and Explanatory Energy

When Latin 6823 operated inside a living medical environment, its users may not have regarded its different sections as strange neighbours.

The herbal, antidotary, doses, substitutions and synonyms all addressed recurring problems within the same therapeutic world.

Its Production Energy was high but available.

Its Explanatory Energy was low because trained readers already supplied the missing connections.

Centuries later, the codex is divided by modern disciplines:

plant illustrations → art history or botany
medical prose → history of medicine
synonyms → philology
recipes → pharmacy history
frontispieces → intellectual history
custody → library history

The original object remains whole.

Modern interpretation separates it.

We then experience the parts as disconnected and must spend enormous effort reconnecting them.

The manuscript’s Explanatory Energy rises not because the book has changed, but because our institutional shelves have cut through its operating structure.

What Latin 6823 establishes

1. Text, image and medical action could coexist inside one manuscript environment

The codex joins illustrated simples to bodily signs, compounds, doses, substitutions and synonyms.

2. Medical knowledge was already modular

Identification, diagnosis, formulation, measurement, substitution and translation appear as distinct but cooperating functions.

3. A medical image library could be extracted from a richer source system

Masson’s plant images and many animal scenes derive from the corresponding visual layers in or around Latin 6823.

4. The medical tradition was international before it became local

The opening authorities visualise a chain connecting Greek, Arabic, Latin and Salernitan medicine to Manfredus and his readers.

5. Scarcity and linguistic drift were recognised operating problems

The Quid pro quo addresses unavailable ingredients, while the synonym collection addresses unstable names.

6. The medical book could function as a resilient system

It contained a standard formulary, extensions beyond that formulary, quantity controls and substitution rules.

7. The visual and textual corpus could travel beyond its first production centre

The manuscript tradition moved into northern Italian courts, workshops and libraries, producing new specialised forms.

What Latin 6823 does not establish

The manuscript does not prove:

• that Manfredus personally painted every image;
• that every text now bound in the codex belonged to one original production plan;
• that the manuscript was used in an apothecary shop;
• that it was routinely carried to patients;
• that every illustration was drawn directly from nature;
• that the present BnF codex was certainly produced in Pisa, Naples or Salerno;
• that it belonged to the Carrara family before entering the Visconti library;
• that every later image manuscript copied directly from this exact physical volume;
• that the written runtime could operate without trained practitioners.

These coordinates remain open.

The manuscript is sufficiently important without closing them prematurely.

The source engine

Latin 6823 gives us the first manuscript in this sequence that resembles an integrated medical operating environment.

World supplies matter.
Image stabilises recognition.
Text supplies attributed properties.
Urine offers a bodily signal.
The antidotary supplies compound formulas.
Dose controls quantity.
Quid pro quo repairs scarcity.
Additional recipes extend the canon.
Synonyms repair language drift.
The trained human connects the whole system to the patient.

This is much more than an illustrated herbal.

It is a book in which the medical world has been decomposed into modules and then placed close enough to operate together.

Later manuscripts carry those modules outward.

One preserves the supply universe.

One preserves the pictures.

One turns them into a retrieval interface.

One translates pharmacology into the language of a court.

Latin 6823 shows the larger machine from which such specialisations become possible.

It does not completely fill the historical void.

But it tells us something decisive about its boundary:

The missing centre was probably not one lost book.

Much of the medical architecture already existed across books like this.

What disappeared most completely was the human coordination layer that knew when to open which section, how to match it to real matter, and how to convert the resulting knowledge into safe action on a living body.

That is where the world must now be built next.

The next manuscript should therefore move from the integrated source engine to its most ambitious courtly expansion:

Historia Plantarum, Rome, Biblioteca Casanatense MS 459

There the Manfredus architecture enters the Visconti world, expands beyond nine hundred entries and becomes one of the largest visual encyclopaedias of medieval medicine.