2. The Manuscript Description Element

2.2.1. Origination and material

The <origDate> and <origPlace> elements are specialised forms of the existing <date> and <name> elements, used to indicate specifically date and place of origin respectively. The <origDate> element is a member of the datable class, and may thus also carry the following attributes:

notBefore	specifies the earliest possible date for the information in the associated element.
	Datatype: CDATA
	Values: Normalised dates should be specified in whatever standardised form is defined by the TEI Header
	Default: #IMPLIED
notAfter	specifies the latest possible date for the information in the associated element
	Datatype: CDATA
	Values: Normalised dates should be specified in whatever standardised form is defined by the TEI Header
	Default: #IMPLIED
evidence	specifies the kind of evidence for the dating supplied by the other attributes.
	Datatype: CDATA
	Legal values:	internal the dating is based on evidence internal to the ms (e.g. an inscription). external the dating is based on evidence external to the ms (e.g. a reference to the ms in some other datable context). conjecture the dating is based on conjecture, which may be further discussed within the associated element.
	Default: #IMPLIED

Omission of the attributes notBefore and notAfter, though permitted by the DTD, is deprecated within the Master project.

As a specialisation of the <name> elemenmt, the <origPlace> element has a reg attribute which may be used to supply a regularized form of the place name.

Detailed information about the history of a manuscript should be encoded within the <history> element discussed in section 2.7. History.

The <material> element may be used to tag any specific term used for the material of which a manuscript is composed, wherever the term occurs.

2.2.2. Dimensions

The <dimensions> element is used to specify measurements, and is thus a specialised form of the existing TEI <measure> element.

<dimensions>

contains any kind of dimensional specification.

Attributes include

type indicates which aspect of the object is being measured. For manuscripts, example values might include: Datatype: CDATA Legal values: leaves dimensions relate to one or more leaves (e.g. a single leaf, a gathering, or a separately bound part) ruled dimensions relate to the area of a leaf which has been ruled in preparation for writing. pricked dimensions relate to the area of a leaf which has been pricked out in preparation for ruling (used where this differs significantly from the ruled area, or where the ruling is not measurable). written dimensions relate to the area of a leaf which has been written, with the height measured from the top of the minims on the top line of writing, to the bottom of the minims on the bottom line of writing. miniatures dimensions relate to the miniatures within the manuscript binding dimensions relate to the binding in which the codex or manuscript is contained box dimensions relate to the box or other container in which the manuscript is stored. Default: #IMPLIED

The <dimensions> may contain a simple description of the measurements which have been made:

<dimensions type="leaves">
Most leaves are 140 x 100 mm. but the outer gatherings are about
5 mm. taller.</dimensions>

Where an institution-wide convention has been established regarding the order in which measurements are to be supplied, their units, etc. this element may contain a simple formula like the following:

<dimensions>12 x 10 x 3</dimensions>

For portability, however, and where conventions differ, it may be preferable to specify the units used, and it may also be useful to specify exactly the dimension along which a measurement has been made. For these reasons, the following three elements are available within the <dimensions> element as an alternative to character data:

<height>	contains a measurement measured along the axis parallel to the spine.
<width>	contains a measurement measured along the axis perpendicular to the spine.
<depth>	contains a measurement measured across the spine.

These elements are all members of the measured class, and thus all carry the following attributes:

units	specifies the units used for this measurement.
	Datatype: CDATA
	Values: The default, and recommended, value is mm, i.e. millimetres.
	Default: #IMPLIED
scope	specifies the applicability of this measurement to the containing manuscript or manuscript part.
	Datatype: (all|most|range)
	Legal values:	all measurement applies to all instances in this manuscript or manuscript part. most measurement applies to most instances in this manuscript or manuscript part. range all the instances measured in this manuscript or manuscript part are within the range specified.
	Default: #IMPLIED

These elements may be given in any order, and may be repeated as often as necessary, with appropriate values for their attributes to indicate the nature and scope of the measurement concerned. For example, in the following case the leaf size and ruled space of the leaves of the manuscript are specified:

<dimensions type="leaves">
 <height scope="range">157-160</height>
 <width>105</width>
</dimensions>
<dimensions type="ruled">
 <height scope="most">90</height>
 <width scope="most">48</width>
</dimensions>

Dimensions of the manuscript being described should normally be specified within the <extent> element (part of the <physDesc> element discussed in 2.6.1. Physical form of the whole ms) rather than elsewhere. Dimensions of specific parts of a manuscript, such as accompanying materials, binding, etc. may however be given in other parts of the description, as appropriate.

2.2.3. References to manuscript locations

The <locus> element is a specialized form of the <ref> element.

<locus>

defines a location within a manuscript or manuscript part, usually as a (possibly discontinuous) sequence of folio references.

Attributes include

from specifies the starting point of the location in a normalised form Datatype: CDATA Values: typically this will be a page number Default: #IMPLIED to specifies the end-point of the location in a normalized form Datatype: CDATA Values: typically this will be a page number Default: #IMPLIED scheme identifies the foliation scheme in terms of which the location is being specified Datatype: CDATA Values: Identifies a foliation scheme, as further defined in the <foliation> element Default: #IMPLIED targets supplies identifiers of associated page images Datatype: IDREFS Values: a list of SGML identifiers, either of <figure> elements within a transcript containing digitized page images, or of <xptr> elements linked to such images Default: #IMPLIED

The <locus> element is used to specify the location in the manuscript occupied by the element within which it appears. If, for example, it is used as the first component of a <msItem> element, or of any of the more specific elements appearing within one (see further section 2.5. Intellectual Content below) then it is understood to specify the location of that item within the manuscript being catalogued.

A <locus> element may be used to identify any reference to one or more folios within a manuscript, wherever such a reference is appropriate.

Locations are conventionally specified as a sequence of folio or page numbers, but may also be a discontinuous list, or a combination of the two. This specification should be given as the content of the <locus> element, using the conventions appropriate to the cataloguing institution, as in the following example:

<msItem n="1"><locus>ff. 1r-24r</locus>
<title>Apocalypsis beati Ioannis Apostoli</title>
</msItem>

A normalised form of the location may also be supplied, using special purpose attributes on the <locus> element, as in the following revision of the above example:

<msItem n="1"><locus from="1r" to="24r">ff. 1r-24r</locus>
<title>Apocalypsis beati Ioannis Apostoli</title>
</msItem>

The targets attribute may be used to associate a <locus> element with one or more other SGML elements in the current document, which should contain either a transcription of the location indicated, or page images of it, as in the following example:

<decoration>
<p>Most of the main body of the book (<locus targets="S1">up to
fol.182v</locus>) was painted and decorated in one style, having links
in style and iconography with the school of Ma&amp;icirc;tre Francois,
although several of the miniatures in this section have been damaged
and overpainted at a later date (e.g. the figure of Christ on <locus
targets="F33R">fol. 33r</locus>; the face of the Shepherdess on <locus
targets="F59V">fol. 59v</locus>, etc.).</p>
</decoration>

Where more than one foliation has been applied to a manuscript, the <locus> element may specify the folio number/s applicable in each foliation used, as in the following example:

<!-- to be supplied -->

2.2.4. Names of persons, places, and organizations

The standard TEI <name> element has attributes type, reg and key, which are used to indicate the type of name (e.g. personal name, placename etc.), to provide a regularised form of the name, and to provide an identifying code or key for it respectively. In addition to these standard TEI facilities, the MASTER DTD also defines an additional role attribute which may be used to specify more exactly the role of a person named in a ms description.

Here are some examples of the use of the <name> element:

<name type="place">Villingaholt</name>
<name type="person" role="scribe">Hoccleve</name>
<name type="person" role="scribe" key="HOC001">Hoccleve</name>
<name type="org" role="owner" reg="Koninklijke Biblioteek">Royal Library</name>

Note that the <name> element is defined as providing information about a name, not a person (or place). In the last example above, the key attribute is used to associate the name with a more detailed description of the person named, provided elsewhere. This more detailed information about a person is provided using the standard TEI element <person>.

<person>

provides structured biographical and other information about any person referenced in a manuscript description.

Attributes include

role specifies the chief role of the person (e.g. as author, scribe, owner etc.) Datatype: CDATA Values: a set of keywords to be defined Default: #IMPLIED sex specifies the sex of the person. Datatype: (m | f | u ) Legal values: m male f female u unknown or inapplicable Default: "m" age specifies the age group to which the participant belongs. Datatype: CDATA Values: suggested values are to be supplied Default: #IMPLIED

For the Hoccleve example given above to be valid, a <person> element must be provided elsewhere which has the value HOCCL1 for its id attribute; the same value will be used as the key attribute of every reference to Hoccleve in the manuscript descriptions (however spelled), but there will only be one <person> element with this identifier.

All the <person> elements referenced by a particular document set should be collected together within a <listPerson> element, located elsewhere in the TEI Header. This functions as a kind of prosopography for all the people referenced by the set of manuscripts being described, in much the same way as a <listBibl> element in the back matter may be used to hold bibliographic information for all the works referenced. (This element is the same as the standard <particDesc> or participant description element, but has been renamed within the Master DTD).

As used in the Master DTD, the <person> element may contain simply one or more paragraphs of running prose, as in the following example:-

  <person id="HOCC01" sex="m" role="scribe">
  <p>Robert Hoccleve, born c.1368 in London, died in 
  Southwick c. 1450. Poet and tragedian, he lived in
  Chester, and was for many years a  Clerk, in the
  office of the Privy Seal</p>
  </person>

More usually however, individual parts of the information about a person will be tagged using the following more specific elements:

<persName>	contains a proper noun or proper-noun phrase referring to a person, possibly including any or all of the person's forename, surname, honorific, added names, etc. (specification)
<birth>	contains information about a person's birth, in particular its date and place.
<death>	contains information about a person's death, notably its date and place.
<residence>	describes a person's present or past places of residence (specification).
<occupation>	contains an informal description of a person's trade, profession or occupation (specification).

The names used for the person should be given first, tagged with the <persName> element, followed by a single <birth> and <death> element, and then any number of <residence> or <occupation> elements, one for each residence, or occupation recorded for the person being described. Each element should use the additional attributes provided by its membership in the datable class, as discussed above (2.2.1. Origination and material); this is the preferred method of dating events such as birth and death of a person. If an exact date is known for a birth or death, it should be supplied on the date attribute. In this case, the notBefore and notAfter attributes need not be supplied, since all attributes are then assumed to have the same value. Dates may also be tagged within the body of these elements, as may names of places (using the <placeName> element).

Here is more structured version of the previous example:

  <person id="HOCC01" sex="m">
  <persName>Hoccleve, Robert</persName>
  <birth notBefore="1365" notAfter="1370">
     <date>c.1368</date>
     <name type="place">London</name>
  </birth>
  <occupation>Clerk, office of Privy Seal</occupation>
  <death notBefore="1440" notAfter="1460">
     <date>c.1450</date>
     <name type="place">Southwick, England</name>
  </death>
  <residence notBefore="1370" notAfter="1460">
     <name type="place">Chester</name>
  </residence>
  </person>

Note that the <persName> and <placeName> elements are used only within a <person> element. Elsewhere, Master recommended practice is to use the more general <name> element. If a person has several distinct names (for example, a nickname), each different name may be tagged, as in the following example:

<!-- to be supplied -->

2.2.5. Catchwords, signatures, secundo folio

The <secFol> element is used to discuss any `secundo folio' information recorded for a manuscript as in the following example:

<secFol>(ando-)ssene in una villa</secFol>

2.2.6. Heraldry

Heraldic descriptions may appear at various points in the description of Western European early modern and medieval manuscripts, usually in the context of ownership information, binding descriptions, or detailed accounts of illustrations. If a description contains a detailed account of the heraldic components of a manuscript independently considered, this should appear as a distinct paragraph within the appropriate section of the physical description. More usually, however, heraldic descriptions will be cited as short phrases within other parts of the record. The phrase level element <heraldry> is provided to enable the cataloguer to mark such phrases for further subsequent analysis, as in the following examples:

Ownership stamp (xvii cent.) on i recto with the arms
<heraldry>A bull passant within a bordure bezanty, 
in chief a crescent for difference</heraldry> [Cole], 
crest, and the legend <q>Cole Deum</q>.'

<p>... partie sup. armori&eacute;e, de gauche &agrave; droite:  armes de
France <heraldry>d'azur &agrave; 3 fleurs de lys d'or</heraldry>; 
de Saint-&Eacute;tienne de Dijon <heraldry>de gueules &agrave; 1 
palme d'or, 2 caillous d'or en chef et 1 en pointe</heraldry>; 
de Cl&eacute;ment VII, antipape <heraldry>&eacute;quipoll&eacute; 
d'or et d'azur</heraldry>, surmont&eacute;es des clefs du 
Saint-Si&egrave;ge; de Philippe le Hardi ou de Bourgogne-Valois 
<heraldry>&eacute;cartel&eacute; au 1er et 4e d'azur
&agrave; la fleur de lys d'or &agrave; la bordure compon&eacute;e d'argent
et de gueules, aux 2e et 3e, band&eacute; d'or et d'azur de six
pi&egrave;ces &agrave; la bordure de gueules<heraldry>; non 
identifi&eacute;es <heraldry>de gueules &agrave; la bande d'or</heraldry>
</p>

2.2.7. Formal definitions

The additional phrase level elements are formally defined as follows:

<!ELEMENT dimensions  %mm;    ((#PCDATA|height|width|depth)*)>
<!ATTLIST dimensions        %a.global;
                               %a.measured;
                               type CDATA #IMPLIED               >
<!ELEMENT height      %mm;   (#PCDATA)   >
<!ATTLIST height            %a.global;
                               %a.measured;               >
<!ELEMENT depth      %mm;   (#PCDATA)   >
<!ATTLIST depth            %a.global;
                               %a.measured;               >
<!ELEMENT width      %mm;   (#PCDATA)   >
<!ATTLIST width            %a.global;
                               %a.measured;               >
<!ELEMENT locus       %mm;  (#PCDATA) >
<!ATTLIST locus    %a.global;
                      from  CDATA #IMPLIED
                      to    CDATA #IMPLIED
                      targets IDREFS #IMPLIED      >
<!ELEMENT origDate %mm;  (#PCDATA)>
<!ATTLIST origDate    %a.global;    
                         %a.datable;  >
<!ELEMENT origPlace %mm; (#PCDATA)>
<!ATTLIST origPlace    %a.global;
                          reg CDATA #IMPLIED      >

<!ELEMENT name %mm; (%phrase.seq;)>
<!ATTLIST name    %a.global;
                     type (person|female|place|org|other) "person"  
                     role CDATA #IMPLIED
                     reg  CDATA #IMPLIED
                     key  IDREF #IMPLIED   >

<!ELEMENT material %mm; (%phrase.seq;)>
<!ATTLIST material    %a.global;      >

<!ELEMENT heraldry %mm; (%phrase.seq;)>
<!ATTLIST heraldry    %a.global;      >

<!ELEMENT catchwords %mm; (%phrase.seq;)>
<!ATTLIST catchwords    %a.global;      >
<!ELEMENT signatures %mm; (%phrase.seq;)>
<!ATTLIST signatures    %a.global;      >
<!ELEMENT secFol %mm; (%phrase.seq;)>
<!ATTLIST secFol    %a.global;      >

The additional elements used within the <listPerson> element in the TEI Header are defined as follows:

<!ELEMENT person %mm; (p+ | (persName*, birth?, death?, 
                               (occupation|residence|bibl)* ))>
<!ATTLIST person %a.global;
                   role CDATA #IMPLIED
                   sex  (m|f|u) "m"
                   age  CDATA #IMPLIED >
<!ELEMENT birth %mm (#PCDATA|date|placeName|ptr|ref|note)* >
<!ATTLIST birth %a.global;
                   date CDATA #IMPLIED
                   %a.datable;       >
<!ELEMENT death %mm (#PCDATA|date|placeName|ptr|ref|note)* >
<!ATTLIST death %a.global;
                   date CDATA #IMPLIED
                   %a.datable;       >
<!ELEMENT residence %mm (#PCDATA|date|placeName|ptr|ref|note)*  >
<!ATTLIST residence %a.global;
                   %a.datable;       >
<!ELEMENT occupation %mm (#PCDATA|date|placeName|ptr|ref|note)*  >
<!ATTLIST occupation %a.global;
                   %a.datable;       >

2.4.1. Languages and scripts

The <textLang> element should be used to provide information about the languages used within a manuscript. It may take the form of a simple note, as in the following example:

<textLang>Old Bulgarian, written in Cyrillic script.</textLang>

Where, for validation and indexing purposes, it is thought convenient to add keywords identifying the particular languages used, the langKey attribute may be used, as in the following example:

<textLang langKey="OBG">Old Bulgarian, written in Cyrillic script.</textLang>

For this example to be valid, the identifier OBG must also be predefined as the value of the ident attribute on a <language> element in the TEI header associated with this description.

<langUsage>
<language id="OBG">Old Bulgarian, written in Cyrillic script.</language>
<!-- other languages used or referenced in the manuscript description -->
</langUsage>

A manuscript will often contain material in more than one language. The langKey attribute should be used only for the chief language. Other languages used may be specified using the otherlangs attribute as in the following example:

<textLang langKey="OBG" otherlangs="RUS HEL">
Mostly Old Bulgarian, with some Russian and Greek material</textLang>

Since Old Bulgarian may be written in either Cyrillic or Glagolitic scripts, and even occasionally in both within the same manuscript, it might be preferable to define identifiers which make the distinction explicit. In such a case, the following <textLang> element might be preferred:

<textLang otherLangs="OBG-CYR OBG-GLA">Old Bulgarian, written in 
Cyrillic and Glagolitic scripts.</textLang>

<langUsage>
<language id="OBG-CYR">Old Bulgarian, written in Cyrillic script.</language>
<language id="OBG-GLA">Old Bulgarian, written in Glagolitic script.</language>
<!-- other languages used or referenced in the manuscript description -->
</langUsage>

Note that the <language> element defines a particular combination of human language and writing system. Only one <language> element may be supplied for each such combination. Standard TEI practice also allows this element to be referenced by any element using the global lang attribute in order to specify the language applicable to the content of that element. For example, assuming that <language> elements have been defined with the identifiers FRA (for French), LAT (for Latin), and DEU (for German), a manuscript description written in French which specifies that a particular manuscript contains German material, might have a <textLang> element like the following:

<textLang lang="fra" langKey="DEU">allemand/latin
</textLang>

2.4.2. Formal definitions

The <msHeading> element is formally defined as follows:

<!ELEMENT msHeading %mm; (#PCDATA|author|respStmt|title| 
                           origPlace|origDate|textLang|note)*>
<!ATTLIST msHeading    %a.global;      >

<!ELEMENT author  %mm; (%phrase.seq;)>
<!ATTLIST author    %a.global;
                       %a.names;
                       attested (yes|no|unk) "unk"
                       accepted (a|y|n|u) "a"              >

<!ELEMENT textLang %mm;  (%phrase.seq;)>
<!ATTLIST textLang    %a.global;
                         langKey IDREF      #IMPLIED     
                         otherLangs IDREFS  #IMPLIED     >

2.5.1. The msItem element

Any combination of the elements described in this section may be used to record information about individual items within the intellectual content of a manuscript or manuscript part. Each discrete item should be encoded within a distinct <msItem> element, and may be classified using the class attribute.

The <msItem> element may not contain any untagged running text. Instead, it may contain any number of any of the following elements, given in any order:

<author>	identifies the primary author of the work or works contained in a manuscript.
Attributes include	attested indicates whether the attribution is explicitly attested in the manuscript. Datatype: yes|no|unk Legal values: yes the attribution is made on the basis of a claim in the ms. added the attribution is made on the basis of a claim added by a subsequent hand in the ms. no the attribution is made on the basis of other evidence. unk the attribution may or may not be attested in the ms. Default: unk accepted indicates whether or not the authorship attribution is generally accepted. Datatype: a|y|n|u Legal values: a the attribution is generally accepted as an attribution. y the attribution is generally accepted to be true. n the attribution is generally accepted to be false. u no claim is made concerning truth of the attribution. Default: a
<bibl>	contains a conventional bibliographic description, for example of a modern edition of the item (standard TEI element)
<colophon>	contains the text of any colophon attached to a particular manuscript item: that is, an inscription, usually found at a break point such as the end of a text or codex, usually containing information about the production of the manuscript, such as the name of the scribe, the date and place of the copying, the person who commissioned the copying, etc.
<decoNote>	contains a note describing either a decorative component of a manuscript, or a fairly homogenous class of such components.; see further 2.6.3. Decoration and binding
<explicit>	contains the text of any explicit attached to a particular manuscript item, that is, the closing words of a text or a section of a text, sometimes used as a kind of title, possibly followed by one or more rubrics or colophons.
Attributes include	type specifies the type of explicit, e.g. whether it is a formal closing for the work. Datatype: CDATA Values: A typology has yet to be defined Default: #IMPLIED defective indicates whether the explicit as given is defective, i.e. incomplete. Datatype: (yes|no|unk) Values: Yes, No, or Unknown Default: no
<incipit>	contains the text of any incipit attached to a particular manuscript item, that is the opening words of a text, frequently used as a form of identifier for it; it may be preceded by one or more rubrics, and may be defective.
Attributes include	type specifies the type of incipit, e.g. whether it introduces a work, is biblical, legal, etc. Datatype: CDATA Values: A typology has yet to be defined Default: #IMPLIED defective indicates whether the incipit as given is defective, i.e. incomplete. Datatype: (yes|no|unk) Values: Yes, No, or Unknown Default: no
<listBibl>	contains a list of related bibliographic descriptions (standard TEI element)
<msitem>	describes an individual work or item within the intellectual content of a manuscript or manuscript part.
Attributes include	class identifies the text types or classifications applicable to this item Datatype: IDREFS Values: One or more codes, each of which is used as the identifier for a text classification element supplied in the TEI Header <textClass> element. Default: #IMPLIED defective indicates whether the item being described is defective, i.e. incomplete. Datatype: (yes|no|unk) Values: Yes, No, or Unknown Default: no
<note>	contains a note or other comment relating to the item (standard TEI element)
<q>	contains any text quoted from the item (standard TEI element)
<respStmt>	supplies a name and description for anyone other than an author credited with intellectual responsibility for some aspect of the item (e.g. an illustrator) (standard TEI element)
<rubric>	contains the text of any rubric or heading attached to a particular manuscript item, that is, a string of words whereby a manuscript signals a text division (e.g. beginning, book, chapter, end) which is in some way set off from the text itself, usually in red ink,or by use of different size or type of script, lining, or other such visual device. .
Attributes include	type specifies the type of rubric, e.g. whether it is at the start or end of the item. Datatype: CDATA Values: A typology has yet to be defined Default: #IMPLIED
<summary>	contains a brief summary of the contents of an item provided by the cataloguer
<textLang>	describes the languages and writing systems used by a manuscript (as opposed to its description, which is described in the langUsage element)
<title>	supplies a title of some kind used for the item (standard TEI element)

The <title> element should be used to supply a regularized form of the item's title, quite distinct from any rubric or incipit quoted from the manuscript. If the item concerned has a standardized or `uniform' title, then this should always be the form preferred as content of the <title> element. If no uniform title exists, or none has been yet identified, then the type attribute on the <title> should be given the value supplied. It is conventional to enclose such supplied titles within square braces, but this will normally be done by a stylesheet. Abbreviated `titles' such as `IMEV 3747' may be tagged using the standard TEI <ref> element, optionally including a pointer to a fuller bibliographic description in a bibliography elsewhere, as further discussed in section 2.8.1.1. Record History.

If supplied, the <author> element should generally contain the normalised form of an author's name, irrespective of how (or whether) this form of the name is cited in the manuscript. If it is desired to retain the form of the author's name as given in the manuscript, this may be tagged as a distinct <name> element, nested within the <author> element with the normalized form of the name on its reg attribute. Alternatively, the normalized form of the name may be supplied as the value of a reg attribute on the <author> element. If value is supplied for either reg or key attributes, then the content of the <author> element is assumed to be a standardized form of name.

Note that the key attribute should be used, as on names in general, to specify the identifier of a <person> element carrying full details of the person concerned (see further 2.2.4. Names of persons, places, and organizations).

Each element within <msItem> has the same substructure, containing any mixture of phrase-level elements and plain text. If a <locus> element is included, in order to specify the location of the component, then it should be given at the start of that element, as in the following example:

<msContents>
<msItem><locus>f. 1-223</locus>
<author>Radulphus Flaviacensis</author>
<title>Expositio super Leviticum </title>
<note>cf. <bibl>Stegm&uuml;ller, RB 7093</bibl></note>
<incipit><locus>f. 1</locus>
Forte Hervei monachi</incipit>
<explicit><locus>f. 223v</locus>
Benedictio salis et aquae</explicit>
</msItem>
</msContents>

In the following example, standard TEI editorial elements have been used to mark the transcription of abbreviations etc. present in the original:

<msItem>
<locus>ff. 1r-24v</locus>
<title>Agrip af Noregs konunga s&ouml;gum</title>
<incipit>regi oc h<expan>ann</expan> setiho
<gap reason="illegible" extent="7"/>sc 
heim se<expan>m</expan> &thorn;io</incipit>
<explicit>h<expan>on</expan> hev<expan>er</expan> 
<expan>oc</expan>&thorn;a buit hesta .ij. aNan vi&thorn; 
f&eacute; enh<expan>on</expan>o<expan>m</expan> aNan til
rei&thorn;<expan>ar</expan></explicit>
<textLang  langKey="ONI">Old Norse/Icelandic</textLang>

As indicated above, a manuscript item may itself contain further nested items, for example where a title or description is supplied for a group of works each of which also has its own rubric, as in the following example:

<msItem>
<locus>ff. 17v, 21v, 34</locus>
<title>Charms numbered 1-3, consisting of
variously formed crosses with inscribed circles</title>

<msItem><locus>f. 17v</locus>
<rubric>Contra inimicos, 1</rubric>
<q>Si quis hoc signum super se portat nequid capi ab
inimico;<./q>
</msItem>

<msItem><locus>f. 21v</locus>
<rubric>Contra mortem subitam, 2</rubric>
<q>Qui hoc signum super se portat sine confessione non
morietur;</q></msItem>

<msItem><locus>f. 34</locus>
<rubric>Pro victoria, 3</rubric>
<q>Hoc signum misit deus Regi Tedeon; qui cum isto pugnat 
victoriam habebit;</q></msItem>

</msItem>

One or more text classification or text-type codes may be specified, either for the whole of a <msContents> element, or for one or more of its constituent <msItem> elements, using the class attribute as specified above. This attribute should specify the identifier used for the appropriate classification within the standard <classCode> element in the TEI Header, as in the following example:

<!-- to be supplied -->

2.5.2. Formal definition

The <msContents> element and its constituents are formally defined as follows:

<!ELEMENT msContents %mm;  (p+| (overview?, msItem+)) >
<!ATTLIST msContents    %a.global; 
         class IDREFS #IMPLIED      
         defective (yes|no|unk) "no"  >
<!ELEMENT msItem %mm;    (locus?, ( author 
                             | respStmt | title |rubric | summary
                             |incipit | explicit | colophon
                             | textLang | q | decoNote 
                             |bibl | listBibl | note  | msItem )* )>
<!ATTLIST msItem    %a.global;
         class IDREFS #IMPLIED
         defective (yes|no|unk) "no"    >
<!ELEMENT summary          %mm;  (%phrase.seq;) >
<!ATTLIST summary    %a.global;      >
<!ELEMENT rubric            %mm;  (%phrase.seq;) >
<!ATTLIST rubric    %a.global;      
       type CDATA  #IMPLIED >
<!ELEMENT incipit           %mm;  (%phrase.seq;) >
<!ATTLIST incipit    %a.global;
       type CDATA  #IMPLIED     
       defective (yes|no|unk) "no"  >
<!ELEMENT explicit          %mm;  ( %phrase.seq;) >
<!ATTLIST explicit    %a.global;
       type CDATA  #IMPLIED       
       defective (yes|no|unk) "no">
<!ELEMENT colophon          %mm;  ( %phrase.seq;) >
<!ATTLIST colophon    %a.global;      >

2.6.1. Physical form of the whole ms

The following elements should appear at the start of a structured physical description in the order specified. They describe various aspects relating to the format of the manuscript object, the material of which it is composed, its size, and physical organization as a sequence of leaves:

<form>	contains a description of the physical format of the manuscript.
<support>	contains a description of the written materials making up the manuscript item.
<extent>	describes the approximate size of the manuscript, specified in any convenient units.(standard TEI element)
<collation>	contains a description of how the leaves or bifolia are physically arranged.

The <form> element may often contain only a single word or phrase: at a later revision of this standard, it is possible that it will be supplied as an attribute value. The <support> element, by contrast, may contain a detailed prose description of the material on which a manuscript is written. Phrase-level elements may be used to tag specific terms of interest if so desired: as in the following example the <material>, <ref>, and <term> elements have been used for this purpose:

<form><p>codex</p></form>
<support>
<p>Mostly <material>paper</material>, with watermarks 
<term>unicorn</term> (<ref>Briquet 9993</ref>) and 
<term>ox</term> (close to <ref>Briquet 2785</ref). 
The first and last leaf of each quire, with the 
exception of quires xvi and xviii, are constituted by
bifolia of <material>parchment</material>, and all 
seven miniatures have been painted on inserted 
singletons of <material>parchment</material>.</p> 
</support> 

If so desired, more specific paragraph-level elements may be used to divide the components of the <support> element. Two such elements are provided:

<overview>	contains the introductory section of any specialised component of a manuscript description.
<watermarks>	contains a detailed description of the watermarks identified in the paper of which a manuscript is composed.

The <overview> element may be used to enclose a general discussion of the support, typically preceding a more detailed discussion, as in the following example:

<!-- to be supplied -->

The <watermarks> element may be used to enclose discussion of one or more watermarks; if this makes reference to standard catalogues of such items, these may be tagged using the standard <ref> element as in the following example:

<support><material>Paper</material>
<watermarks>Watermark: <term>anchor in a
circle with star on top</term>, 
<term>countermark B-B with trefoil</term> 
similar to <ref>Moschin, Anchor N 1680</ref> 
<date>1570-1585</date>.</watermarks></support>

The <extent> element is principally a specification of the number of leaves or bifolia a manuscript contains, but may also contain any number of other measurements carried out on the manuscript. Such measurements may be specifically marked using the phrase level <dimensions> element, as in the following example, or left as plain prose.

<extent>2 leaves, taken from the binding of a printed book</extent> 
<extent>ii+321 leaves</extent>
<extent>37+212 leaves</extent>

The <collation> element should be used to specify exactly how the leaves or bifolia of a manuscript are combined into quires etc. This may be described using informal prose, or any appropriate notational convention. No specific notation is defined here as yet, but provision is made for the tagging of such formulæ with the standard TEI <formula> element as in the following example:

<collation>
<p>
<formula>1-5.8 6.6 (catchword, f. 46, does not match following text) 
7-8.8 9.10, 11.2 (through f. 82) 12-14.8 15.8(-7)</formula>
<catchwords>Catchwords are written horizontally in center 
or towards the right lower margin in various manners: 
in red ink for quires 1-6 (which are also signed in red 
ink with letters of the alphabet and arabic numerals); 
quires 7-9 in ink of text within yellow decorated frames; 
quire 10 in red decorated frame; quire 12 in ink of text; 
quire 13 with red decorative slashes; quire 14 added in 
cursive hand.</catchwords>
</p></collation>

2.6.2. Writing and Notations

The second group of elements within a structured physical description concerns aspects of the writing or other notation (notably, music) found within a manuscript.

<layout>	contains a description of ruling technique and layout.
Attributes include	columns specifies the number of columns per page Datatype: CDATA Values: either a whole number or the string "NA" (not applicable) Default: '1' ruledLines specifies the number of ruled lines per column Datatype: CDATA Values: may be given as a pair of numbers (a range), as a whole number, or as the string "NA" (not applicable) Default: #IMPLIED writtenLines specifies the number of written lines per colum Datatype: CDATA Values: may be given as a pair of numbers (a range), as a whole number, or as the string "NA" (not applicable) Default: #IMPLIED
<msWriting>	contains a description of all the different kinds of writing used in a manuscript.
Attributes include	hands specifies the number of distinct hands identified within the manuscript Datatype: CDATA Values: A whole number, or the string "many" Default: #IMPLIED
<handDesc>	describes a particular style or hand distinguished within a manuscript.
Attributes include	scribe gives a standard name or other identifier for the scribe believed to be responsible for this hand. Datatype: CDATA Values: 'tremulous hand', 'Hoccleve', etc. Default: #IMPLIED script characterizes the particular script or writing style used by this hand. Datatype: CDATA Values: 'secretary', 'copperplate', 'Chancery', 'Italian', etc. Default: #IMPLIED medium describes the tint or type of ink, e.g. 'brown', or other writing medium, e.g. 'pencil', Datatype: CDATA Default: #IMPLIED scope specifies how widely this hand is used in the manuscript. Datatype: (sole|major|minor) Legal values: sole only this hand is used throughout the manuscript major this hand is used through most of the manuscript minor this hand is used occasionally in the manuscript Default: #IMPLIED
<musicNotation>	contains description of type of musical notation.

<layout ruledLines="25-32">
<p>Most pages have between 25 and 32 long lines ruled in lead.</p>
</layout>

<msWriting>
<p>Written in a <term>late caroline minuscule</term>; versals in a 
form of <term>rustic capitals</term>; although the marginal and 
interlinear gloss is written in varying shades of ink that are 
not those of the main text, text and gloss appear to have been 
copied during approximately the same time span.</p>
</msWriting>

<msWriting hands="2">
<p>The manuscript is written in two contemporary hands, otherwise
unknown, but clearly those of practised scribes.  Hand I writes
ff. 1r-22v and hand II ff. 23 and 24. Some scholars, notably
Verner Dahlerup and Hreinn Benediktsson, have argued for a third hand
on f. 24, but the evidence for this is insubstantial.</p>
</msWriting>

<handDesc><p>Written in insular phase II half-uncial
with interlinear Old English gloss in an Anglo-Saxon 
pointed minuscule.</p></handDesc>

<musicNotation>
<p>Neumes in campo aperto of the St. Gall type.</p>
</musicNotation>

2.6.3. Decoration and binding

In describing a manuscript, it is often difficult or impossible to draw a clear distinction between aspects which are purely physical and aspects which contribute to the intellectual content. This is particularly true of `decorative' aspects such as illustrations and decorations within the manuscript or binding. We propose the following elements for the purpose of delimiting discussion of these aspects within a manuscript description, amd for convenience locate them all within the physical description, despite the fact that illustrative features in many cases may be better thought of as part of the intellectual content, and thus better described under the <msContents> element discussed in section 2.5. Intellectual Content.

<decoration>	contains a description of the decoration of a manuscript, either as a sequence of paragraphs, or as a sequence of topically organised decoNote elements.
<decoNote>	contains a note describing either a decorative component of a manuscript, or a fairly homogenous class of such components.
Attributes include	type specifies the kind of decorative feature being described Datatype: CDATA Values: Any convenient typology may be used Default: #IMPLIED subtype supplies a further sub-categorization of the value specified by the type attribute. Datatype: CDATA Values: Any convenient typology may be used Default: #IMPLIED figurative indicates whether the note concerns purely figurative aspects of the decoration Datatype: (yes|no|na) Legal values: yes figurative features are discussed no figurative features are not discussed na not applicable Default: 'na' technique the note primarily concerns the technique or style of the decorative features Datatype: CDATA Values: Any convenient typology may be used Default: #IMPLIED quality the note contains an assessment of the aesthetic quality of the decorative feature Datatype: CDATA Values: Any convenient typology may be used Default: #IMPLIED figurative indicates whether the note concerns purely figurative aspects of the decoration Datatype: (yes|no|na) Legal values: yes figurative features are discussed no figurative features are not discussed na not applicable Default: 'na' illustrative indicates whether the note concerns purely illustrative aspects of the decoration Datatype: (y|n|u) Legal values: y illustrative features are discussed n illustrative features are not discussed u unknown or inapplicable Default: 'u'
<bindingDesc>	describes the present and former bindings of a manuscript, either as a series of paragraphs or as a series of distinct binding elements, one for each binding of the manuscript.
<binding>	contains a description of one binding, i.e. type of covering, boards, etc. applied to a manuscript
Attributes include	contemporary specifies whether or not the binding is contemporary with the majority of its contents Datatype: (yes|no|unk) Legal values: yes the binding is contemporaneous with its contents no the binding is not contemporaneous with its contents unk the date of either binding or manuscript is unknown Default: 'unk'

<decoration>
<p>The decoration comprises two
full page miniatures, perhaps added by the original
owner, or slightly later; the original major decoration
consists of twenty-three large miniatures, illustrating
the divisions of the Passion narrative and the start of
the major texts, and the major divisions of the Hours;
seventeen smaller miniatures, illustrating the suffrages
to saints; and seven historiated initials, illustrating
the pericopes and major prayers.</p>
</decoration>

<decoration>
<decoNote type="miniature" technique="painted" figurative="yes">
<p>One full-page miniature, facing the beginning of the first Penitential Psalm.</p></decoNote> 
<decoNote type="initial" technique="painted" figurative="yes">
<p>One seven-line historiated initial, commencing the first Penitential Psalm.</p>
</decoNote>
<decoNote type="initial" technique="painted" figurative="no">
<p>Six four-line decorated initials, commencing the second through the 
seventh Penitential Psalm.</p></decoNote>
<decoNote type="initial" technique="pen-flourished" figurative="no">
<p>Some three hundred two-line versal initials with pen-flourishes, 
commencing the psalm verses.</p></decoNote>
<decoNote type="border" technique="painted" figurative="no">
<p>Four-sided border decoration surrounding the miniatures and 
three-sided border decoration accompanying the historiated and 
decorated initals.</p></decoNote></decoration>

<decoration>
<decoNote type="miniatures">
<p>Fourteen large miniatures with arched
tops, above five lines of text:
<list>
<item><locus>fol. 14r</locus>Pericopes. <term>St. John writing on
Patmos</term>, with the Eagle holding his ink-pot and pen-case;
some flaking of pigment, especially in the sky</item>
<item><locus>fol. 26r</locus>Hours of the Virgin, Matins.
<term>Annunciation</term>; Gabriel and the Dove to the
right</item>
<item><locus>fol. 60r</locus>Prime. <term>Nativity</term>; the
<term>Virgin and Joseph adoring the Child</term></item>
<item><locus>fol. 66r</locus>Terce. <term>Annunciation to the
Shepherds</term>, one with <term>bagpipes</term></item>
<!-- ... -->
</list></p>
</decoNote>
</decoration>

<bindingDesc>
<p>Sewing not visible; tightly rebound over
19th-cent. pasteboards, reusing panels of 16th-cent. brown leather with
gilt tooling &agrave; la fanfare, Paris c. 1580-90, the centre of each
cover inlaid with a 17th-cent. oval medallion of red morocco tooled in
gilt (perhaps replacing the identifying mark of a previous owner); the
spine similarly tooled, without raised bands or title-piece; coloured
endbands; the edges of the leaves and boards gilt.Boxed.</p>
</bindingDesc>

<binding>Bound, s. XVIII (?), in 
<material>diced russia leather</material>
retaining most of the original 15th century 
metal ornaments (but with some replacements) 
as well as the heavy wooden boards; 
<decoNote>on each cover: alternating circular stamps of the 
Holy Monogram, a sunburst, and a flower;</decoNote> 
<decoNote>on the cornerpieces, one of which is missing,
a rectangular stamp of the Agnus Dei;</decoNote> 
rebacked during the 19th (?) century.
</binding>

2.6.4. Foliation, additions, and condition

The following elements are used to describe the remaining subdivisions of the physical description:

<foliation>	describes the numbering system or systems used to count the leaves or pages in a codex.
<condition>	contains a description of the physical condition of the manuscript.
<additions>	contains a description of any significant additions found within a manuscript, such as marginalia or other annotations.

<foliation>
<p>Folio numbers were added by &amp;Aacute;rni Magn&amp;uacurte;sson 
ca. <date>1720-1730</date> with brown ink in the upper right corner 
of all recto-pages.</p></foliation>

<foliation><p>Page numbers have been written with pen in the late 
19th century on top of every ten recto-pages, to the right: 11, 
21, 31, etc. Later, folio numbers have been written between 
columns on top of every recto-page, in pencil.</p></foliation> 

<foliation>
<p>Quire and leaf signatures in letters, [b]-v, and roman
numerals; those in quires 10 (1) and 17 (s) in red ink and different
from others; every third quire also signed with red crayon in arabic
numerals in the center lower margin of the first leaf recto: "2" for
quire 4 (f. 19), "3" for quire 7 (f. 43); "4," barely visible, for
quire 10 (f. 65), "5," in a later hand, for quire 13 (f. 89), "6," in
a later hand, for quire 16 (f. 113).</p>
</foliation>

<additions><p>In most parts of the codex the text has been quite
extensively corrected in a contemporary hand from the manuscript GKS
3270 4to.</p>
</additions>

<additions>
<p>The text of this manuscript is not interpolated with
sentences from Royal decrees promulgated in 1294, 1305 and 1314. In
the margins, however, another somewhat later scribe has added the
relevant paragraphs of these decrees, see pp. 8, 24, 44, 47
etc.</p>
<p>As a humorous gesture the scribe in one opening(?) of the
manuscript, pp. 36 and 37, has prolonged the lower stems of one letter
f and five letters  &amp;thorn; and has them drizzle down the
margin.</p>
</addition>

<condition>
<p>The text begins and ends defective, and there are two lacunae toward
the end. What was originally the first leaf of the first quire,
numbered 1a, has been cut away, leaving a strip approximately 15 mm
wide; one to two letters, in some places on the verso up to four,
remain of each line.  The manuscript contains a large number of
erasures, possibly by the scribe, possibly by a later hand, but in any
case for the most part unfilled in; for example on f. 3ra, lines 15-6
there are two erasures, the first of perhaps 16 the second of six
characters, and four lines of text have been erased on f. 5vb
following l. 21.  In addition, there are several spaces that appear
never to have been filled in, e.g. f. 13vb, l. 24, where enough space
for about eight letters has been left blank by the scribe.
</p></condition>

2.6.5. Formal definitions

The formal definition for the <physDesc> element and its constituents is as follows:

<!ELEMENT  physDesc    %mm;   (p+ | (form?, support?, extent?,  
                                 collation?, layout?, msWriting?, musicNotation?, 
                                 decoration?, bindingDesc?, foliation*, 
                                 additions?, condition?))> 
<!ATTLIST  physDesc    %a.global; >
<!ELEMENT  form        %mm;   (p+)                   >
<!ATTLIST  form        %a.global;  >
<!-- extent is defined elsewhere -->
<!ELEMENT  support     %mm;   (p+ | (overview?,watermarks?))                  >
<!ATTLIST  support     %a.global;  >
<!ELEMENT  overview    %mm;   (p+)                   >
<!ATTLIST  overview    %a.global;  >
<!ELEMENT  watermarks  %mm;   (p+)                   >
<!ATTLIST  watermarks  %a.global;  >
<!ELEMENT  collation   %mm; (p+)>
<!ATTLIST  collation        %a.global;  >
<!ELEMENT  layout      %mm;   (p+)       >
<!ATTLIST  layout      %a.global;
                           columns CDATA '1'
                           ruledLines CDATA #IMPLIED     
                           writtenLines CDATA #IMPLIED     >
<!ELEMENT  msWriting      %mm;    ((handDesc|p)+)        >
<!ATTLIST  msWriting     %a.global; 
                            hands CDATA #IMPLIED >
<!ELEMENT  handDesc        %mm;    (p+) >
<!ATTLIST  handDesc        %a.global    
                           scribe             CDATA               #IMPLIED
                           script             CDATA               #IMPLIED
                           medium             CDATA               #IMPLIED
                           scope    	(sole|major|minor)        #IMPLIED >
<!ELEMENT  musicNotation       %mm;    (p+)                   >
<!ATTLIST  musicNotation        %a.global;  >
<!ELEMENT  decoration  %mm;    (p+|decoNote+)        >
<!ATTLIST  decoration        %a.global;  >
<!ELEMENT  decoNote    %mm;    (p+)                 >
<!ATTLIST  decoNote      %a.global;  
                             type CDATA #IMPLIED
                             subtype    CDATA #IMPLIED      
                             size       CDATA #IMPLIED
                             technique  CDATA #IMPLIED 
                             quality    CDATA #IMPLIED 
                             figurative (yes|no|na) 'na'   
                             illustrative (y|n|u)   'u'    >
<!ELEMENT  bindingDesc     %mm; ((p|decoNote)+ | binding+)>
<!ATTLIST  bindingDesc       %a.global;     >
<!ELEMENT  binding     %mm; ((decoNote|p)+)>
<!ATTLIST  binding 
               %a.global;
               %a.datable;         
              contemporary (yes|no|unk)  "unk" >
<!ELEMENT  foliation   %mm; (p+)>
<!ATTLIST  foliation        %a.global;  >
<!ELEMENT  additions  %mm; (p+)  >
<!ATTLIST  additions        %a.global;  >
<!ELEMENT  condition   %mm; (p+)>
<!ATTLIST  condition        %a.global;  >

2.8.1. Administrative Information

A variety of information relating to the curation and management of a manuscript may be recorded as simple prose narrative tagged using the standard <p> element. Alternatively, different aspects of this information may be presented grouped within one or more of the following specialized elements:

<recordHist>	provides information about the source and revision status of the parent manuscript description itself.
<custodialHist>	contains a description of a manuscript's custodial history, either as running prose or as a series of dated custodial events.
<availability>	supplies information about the availability of a manuscript, for example any restrictions on its use or access, its copyright status, etc. (standard TEI element)
<remarks>	contains any comments or remarks not forming part of the description proper, for use by the cataloguer.

<!-- to be supplied-->

<source><p>Directly catalogued from the original manuscript</p></source>

<source><p>Information transcribed from 
<bibl><title>IMEV</title><biblScope>1234</biblScope></bibl>
</p></source>

<additional>
<recordHist>
<source><p>Information transcribed from 
<ref target="IMEV123">IMEV 123</ref>
</p></source>
</recordHist>
<!-- ... -->
<listBibl>
<bibl id="IMEV123">
<title>Index of Medieval Verse</title>
<!-- other bibliographic details for IMEV here -->
<biblScope>123</biblScope>
</bibl>
<!-- other bibliographic records relating to this manuscript here -->
</listBibl>
</additional>

<!-- to be supplied -->

<availability><p>The manuscript is in poor condition, due to many of
the leaves being brittle and fragile and the poor quality of a number
of earlier repairs; it should therefore not be used or lent out until
it has been conserved.</p></availability>

<custodialHist>
<custEvent type="conservation" notBefore="1961-03" notAfter="1963-02">
<p>Conserved between March 1961 and February 1963 at Birgitte Dalls 
Konserveringsv&aelig;rksted.</p></custEvent>
<custEvent type="photography" notBefore="1988-05-01" notAfter="1988-05-30">
<p>Photographed in May 1988 by AMI/FA.</p></custEvent>
<custEvent type="transfer/dispatch" notBefore="1989-11-13" notAfter="1989-11-13" >
<p>Dispatched to Iceland 13 November 1989.</p></custEvent>
</custodialHist>

<remarks>
<p>Re ff. 33-36v:  According to Leroquais, the feast of Corpus
Christi with its octave was instituted for Franciscans in 1319; 
the hymn here is the standard Tantum ergo.</p>

<p> Re a photo:  Yes, do one, since the manuscript is signed and dated! 
How about f. 7v (nice cadelled initials, and f. 18 with Clare 
has already been reproduced elsewhere).</p>
</remarks>

<!ELEMENT  adminInfo   %mm;
              (p+ | (recordHist?, availability?, custodialHist?, remarks?)) >
<!ATTLIST  adminInfo        %a.global;  >
<!ELEMENT  recordHist %mm;  (source, change*)   >
<!ATTLIST  recordHist        %a.global;  >
<!ELEMENT  source %mm; (p+)   >
<!ATTLIST  source        %a.global;  >
<!-- the change element is defined elsewhere -->
<!ELEMENT  custodialHist %mm;  (p+ | custEvent+) >
<!ATTLIST  custodialHist        %a.global;  >
<!ELEMENT  custEvent %mm;  (p+)>
<!ATTLIST  custEvent        %a.global;
                               %a.datable;  
                               type  CDATA #IMPLIED >
<!ELEMENT  remarks %mm;  (p+)>
<!ATTLIST  remarks        %a.global;  >

2.8.2. Surrogates

The <surrogates> element is used to provide information about any digital or photographic representations of the manuscript which may exist within the holding institution or elsewhere.

<surrogates>

contains information about any digital or photographic representations of the manuscript being described which may exist in the holding institution or elsewhere.

The <surrogates> element should not be used to repeat information about representations of the manuscript available within published works; this should normally be documented within the <listBibl> element within the <additional> element. However, it is often also convenient to record information such as negative numbers, digital identifiers etc. for unpublished collections of manuscript images maintained within the holding institution, as well as to provide more detailed descriptive information about the surrogate itself. Such information may be provided as prose paragraphs, within which identifying information about particular surrogates may be presented using the standard TEI <bibl> element, as in the following example:

<surrogates><p>
<bibl>
  <title type="gmd">diapositive</title>
  <idno>AM 74 a, fol.</idno>
   <date>May 1984</date>
</bibl>
<bibl>
  <title type="gmd">b/w prints</title>
  <idno>AM 75 a, fol.</idno>
   <date>1972</date>
</bibl>
</p></surrogates>

At a later revision, the content of the <surrogates> element is likely to be expanded to include elements more specifically intended to provide detailed information such as technical details of the process by which a digital or photographic image was made.

If the whole of a manuscript is being digitized, it should be contained in a separate document which lists all the images using <div> etc. to structure them. In such a case, the <msDescription> placed within the teiHeader.

The <surrogates> element is formally defined as follows:

<!ELEMENT surrogates  %mm;    (p+)>
<!ATTLIST surrogates  %a.global; >

2.8.3. Accompanying material

The circumstance commonly arises where a manuscript has additional material, not originally part of the manuscript, which is bound with it or otherwise accompanying the manuscript. In cases where this is clearly a distinct manuscript, the whole manuscript should be treated as a composite manuscript and the additional matter described in a separate <msPart> (see 2.9. Manuscript Parts below). However, there are cases where the additional matter is not self-evidently a distinct manuscript: it might be an important set of notes by a later scholar or owner, or it might be a file of correspondence relating to the manuscript. The <accMat> element is provided as a holder for this kind of information:

<accMat>	contains details of any significant additional material which may be closely associated with the manuscript being described, such as non-contemporaneous documents or fragments bound in with the manuscript at some earlier historical period.
Attributes include	type further characterizes the accompanying material, for example as letter, note, paste-in, etc. Datatype: CDATA Values: any convenient typology may be used Default: #IMPLIED

Here is an example of the use of this element, describing a note by the Icelandic manuscript collector &Aacute;rni Magn&uacute;sson which has been bound with the manuscript:

<accMat>
<p>A slip in &amp;Aacute;rni Magn&amp;uacute;sson's hand has been stuck to the
pastedown on the inside front cover; the text reads:
<q lang="IS">&amp;THORN;idreks S&amp;oslash;gu &amp;thorn;essa hefi eg 
feiged af Sekreterer Wielandt Anno 1715 
i Kaupmanna h&amp;oslash;fn.  Hun er, sem eg sie, Copia af Austfirda
b&amp;oacute;kinni (Eidag&amp;aacute;s) en<expan>n</expan> ecki progenies
Br&amp;aelig;dratungu bokarinnar. Og er &amp;thorn;ar fyrer eigi i
allan<expan>n</expan> m&amp;aacute;ta samhlioda
&amp;thorn;<expan>eir</expan>re er Sr Jon Erlendz son hefer ritad fyrer
Mag. Bryniolf. &amp;THORN;esse &amp;THORN;idreks Saga mun vera komin fra Sr
Vigfuse &amp;aacute; Helgafelle.</q></p>
</accMat>

The formal definition for the <accMat> element is as follows:

<!ELEMENT accMat %mm;    (p+)>
<!ATTLIST accMat %a.global; 
             type CDATA #IMPLIED>