The Pali Text Society has let A. K. Warder's Pāli Metre fall out of print, and are now offering it in PDF format instead: here.

§30. The standpoint of the present work is more independent of the traditions of Greek scholars than these other contributions, or those of Helmer Smith, have been. There appears to be no reason to suppose that the Indian rhythms had any special resemblance to the rhythms of Greek metres. On the contrary the impression of the present writer is that those Western scholars with a Western Classical education who read, for example, ("Classical") Sanskrit poetry according to the habits of scansion they acquired when studying Greek poetry, thereby destroy the beauty, the variety and especially the syncopations of the Indian rhythms. Indian scholars do not recite Sanskrit poetry in that manner. Their renderings encouraged the present writer to follow his inclination as a music lover (with more training in music, Western and Indian, than in Greek and Latin), fascinated by what appeared to him to be the musical rhythms of, in particular, akṣaracchandases ("fixed syllabic" metres), to take the Sanskrit patterns in a strictly "measured" manner. Instead of reducing them to the regular feet of Greek metrics, through anceps, "drag" and the like, he thus realized an immense variety of different rhythms. If the scansion of all these akṣaracchandases should be limited to a few trochaic, dactylic, etc., patterns as in Greek, why are there so many of them in regular and carefully contrasted use?

(For a less exhaustive overview of the topic, try Ven. Ānandajoti's Outline of the Metres in the Pāḷi Canon.)