flagging divergences between multiple hands of the same manuscript

[matteodifranco]

The second or third hands of the witnesses are encoded as hand as well as wit within the manuscript description, so that:

<witness xml:id="R" corresp="#R.hand.1">R
[...]
    <physDesc>
        <handDesc hands="2">
            <handNote xml:id="R.hand.1">First hand before correction</handNote>
            <handNote xml:id="R.hand.2">Second hand</handNote>
        </handDesc>
    </physDesc>
[...]
    <listWit>
        <witness xml:id="R.2" corresp="#R.hand.2">
            <label>R<hi rend="sup">2</hi></label>
        </witness>
        <witness xml:id="R.3" corresp="#R.hand.3">
            <label>R<hi rend="sup">3</hi></label>
        </witness>
    </listWit>
</witness>

This should be treated in the same way as families that include multiple witnesses.

In the collation visualization, when the second hand differs from the first, only the reading of the first hand is displayed, as in the following example:
R has γῆ, R² has γ’ ἡ as in the text

<app from="#Cyzicus-09-20" to="#Cyzicus-09-21">
    <lem wit="#O #R.2" exclude="#R">γ’ ἡ</lem>
    <rdg wit="#R #V #F #I #K #Pa #VaBo #D #U #Ab #T #Upp #Mq241 #Vat931
        #Iunt #Steph #Jebb #Dindorf" exclude="#R.2" type="associative" cause="diction">γῆ
        <note>add. in mg VaBo</note>
    </rdg>
    <rdg wit="#Vat932" type="lectio.sing" cause="diction">θ’ ἡ</rdg>
    <rdg wit="#Cs" type="lectio.sing" cause="omission"/>
</app>

The screen also displays the issue of lemma repetition noted in #269.