That of No. 32-3 is the best, because the true relation (apposition) of the words is maintained, and is shown by a mark (the comma) already seen to be the proper mark for this relation. The failure of the comma to show the relation of apposition in No. 32 is due to the fact that the words following “officers” form an _apparent_ series with “officers.” The comma here needs re-enforcing by a special grouping mark, which office the dash performs. The comma shows the relation; the dash does the grouping.
We spoke above of extending the office and limiting the value of a mark. We mean by this that the more uses and, in consequence, the more meanings a mark has, the more difficult it is for the reader to interpret it when he meets it. For this reason the comma and dash seem to be the best mode of punctuating sentences like the above.
As we saw in Sentence 6, the colon is used to mark the largest divisions of a sentence; and it is used, as in Sentence 25, for the formal introduction of particulars. The former use is determined by the colon’s rank, which is above the semicolon and below the period; the latter use is conventional, and is without reference to its rank.
If two colons appear in a sentence, one because of its rank and one in the enumeration of particulars, there may be a seeming inconsistency in grouping. The same is true when a colon is used within a group made by a semicolon. For instance, in No. 25-3 we used a colon in each of two groups, the groups themselves being separated by a semicolon. We think it much better to use within each group the comma and dash than to use the colon, thus emphasizing the grouping done by the semicolon, instead of apparently subordinating the colon in each group to the semicolon making the two groups. This would give the following punctuation for No. 25-3:
33. In a bill of exchange there are three original parties,—drawer, drawee, and payee; in a promissory note, only two,—maker and payee.
We have thus used a comma and a dash to set off particulars formally introduced, making the punctuation of No. 33 inconsistent with that of No. 25, and apparently inconsistent with No. 32-2. If this resulted in misleading a reader, objection might be raised; but, we feel confident, this exceptional mode of punctuation is justified.