William Gardner Hale, The Art of Reading Latin: How To Teach It

Up to the fourth week inclusive, the Latin was written upon the board at these weekly exercises, one word at a time, the questions being put, as indicated by the footnotes in the papers given below, at one point and another as the sentence progressed. For several weeks after that time, the Latin sentence was written by each student, one word at a time, as pronounced by the instructor, the questions being set and answered as before. After this, the writing of the Latin was forbidden, and the passages used were interpreted only as heard from the instructor's reading.

At the first interview, the class had worked out, as it was put upon the board, one word at a time, the sentence in Livy I.1.5.

“Ibi egressi Troiani, ut quibus ab immenso prope errore nihil praeter arma et naves superesset, cum praedam ex agris agerent, Latinus rex Aboriginesque, qui tum ea tenebant loca, ad arcendam vim advenarum armati ex urbe atque agris concurrunt.” (1.1.5)

As we reached the point ...ut quibus, they had made out, under questioning, that ut might be (1) a conjunction, in which case quibus could be (a) an interrogative introducing an indirect question depending on the ut-verb, or (b) a relative referring to something connected with the ut-verb; or that, on the other hand, ut might be (2) an adverb, in which case the quibus-clause must be substantially an adjective modifying Troiani; in other words, a characterizing clause. In this connection, they had been told, for the sake of having the whole matter secured for their repertory of combinations of this kind, that what was essential in this latter case was the characterizing clause itself, and that in strictness no introductory word was necessary; if one was used, however, it might be either ut, utpote, or quippe; and it was also pointed out that, while there were three possibilities for a combination like ut quibus, there was only one possibility for a combination like utpote quibus or quippe quibus.

As we reached superesset, it was pointed out, against the practical habit of thought of nearly all the class, that, since in Latin the common practice was to put a modifying clause or phrase before the thing modified, the chances were that the quibus-clause, if it should turn out to be a characterizing clause, would bear, not upon egressi, but upon something which we were still to wait for. (This something turned out to be cum ... agerent, — the natural thing for destitute men to do.)

As we reached ...cum praedam, at which stage it was sure that cum was a conjunction, the point was made, though again against the sentiments of the class, that Troiani was the subject of the verb introduced by cum, since the Romans were fond of taking out a conspicuous word or phrase belonging to an introductory temporal sentence, and putting it before the connective.¹

The passage chosen for the first written exercise turned out to be a little too difficult in the reasoning at the et cui point, though it had a certain and considerable usefulness in displaying to the class a sentence of which some of them, though knowing the meaning of each word, and though able to "parse" it from beginning to end if it were once translated to them, would yet fail to comprehend the meaning, through a lack of a working knowledge of the constructions involved.