The <idx> tags denote an index entry. These should be placed within the element that they describe. By this we mean that an <idx> element can be placed within a <theorem> to refer to just that theorem, or it might be placed within a <subsection> to refer to that subsection. When you do this, a natural place to place the <idx> is right after the <title> and similar metadata. In this way, electronic versions of your work can have an index that is more informative than a traditional index that uses just page numbers, since it will be apparent while reading and using the index just what type of object the entry refers to. (See the end of this Guide in an electronic format for an example, Index). Note that the text contained within the <idx> tags does not actually appear in the article—it only serves to mark the location the index entry points to. You can have several levels of headings by structuring your <idx> element with up to three <h> tags (“h” for “heading”). Additionally, you can use <see> or <seealso> for cross-references within the index.
A similar device is used to create a list of notation for a technical (mathematical) work. Place a <notation> element as close as possible to the place where notation is first introduced. If you use the <definition> tag for your definitions, then this is a very natural place to also introduce notation. Inside of <notation> use the <usage> tag to include a short example of the notation in use, wrapped in an <m> tag and use syntax (as usual). The <description> tag should contain a very short description in words of what the notation is for. So “center of a group” would be a good description to accompany the usage “$$Z(G)\text{.}$$