In 1912 H. Maspero endeavoured to bring some order into the apparently capricious repartition of the old ju-shêng tones among the remaining four tones of Pekingese, and noted that words in plosive initials (including aspirated plosives), and in nasal or liquid initials, tend to be found in the departing tone (tone IV), and to a less extent in the upper even (tone I); while those with otḥer initials gravitate towards the old lower even (tone II).

In 1915 Karlgren considered Maspero's treatment of the problem, and while in general agreeing with his conclusions, pointed out that his law covered the facts only very approximately in regard, e.g., to words in unaspirated plosive initials. Nevertheless he quotes with approval Maspero's explanation (applied also to the twofold development of the lower rising tone in the modern Northern Chinese dialects), viz., that the initials 1-, m-, n-, n-, and vocalic initials involved a pronunciation of the syllable somewhat different from that of words having affricate, occlusive, or fricative initials; and he proceeds to suggest that the difference lay in a somewhat higher pitch in the case of the former type of word. This explanation is, however, difficult to follow; if, e.g., ACh dz‘ua, became \tso (tone IV) rather than (tone III) because the fourth tone is more compatible with the deeper pitch proper to affricate initials, it is by no means clear why ACh dz‘ak, having the same initial, should become ‘tso (tone II) rather than enter tone IV with so many other words from the lower ju-shêng.