We next notice that ten is expressed by two separate words, kèp (in ten and its multiples) and krē (in the compounds from eleven to nineteen). So far as vocabularies have yet been published, the only other tribes of the Tibeto-Burman family[3] which have a word resembling kèp for ten are Maring Naga, one of the Naga-Kuki languages, where it is chip, and Shö or southern Chin, where gip is used for ten in the sequence thirty, forty, fifty, etc. (thum-gip, thirty, mlī-gip, forty, nghā-gip, fifty, exactly corresponding to the Mikir thòm-kèp, philī-kèp, phòngō-kèp). The close resemblance of the other numerals in Maring to Mikir forms is noticeable; four is filī, five fungā, and six tharuk. The other word for ten, krē, strongly resembles the Angāmi kerr or kerru and the Kachcha Naga gārēo; in the Central Naga group the prefix ke- has been replaced by ta- or te-, and the words for ten are ter, tarā (Ao), tarō, tarā, tarē (Lhota), etc. In the Naga-Kuki group Sopvoma has chirō, Marām kero, Tangkhul tharā, etc. In the Kuki-Chin group Meithei tarā is the same word: in the Central Chin another prefix, pu-, pö-, or pā-, is used, and we have Lai pörā, Banjogi parā, Taungthā parhā. There are no Boṛo forms which correspond to krē, nor any much resembling kèp.

The Mikir word for twenty, ingkoi, is made up of the prefix ing-, and koi, formerly (before the loss of the final l) kol. Kachcha Naga has the same word, engkai, Kabui choi, koi, or kol. The word also appears in Garo (kol), Tipura (khol), and Deori-Chutiya (kwa), of the Boṛo group; Angāmi me-kwū, me-khi, mekko, Lhota me-kwī, mekwü, in the West and Central Naga groups; Marām and Sopvoma (makē, makei), Tangkhul (magā), Phadeng (ma-kui), in the Naga-Kuki group, and Singpho khun. In the Kuki-Chin languages it is very common (Meithei kul, Siyin kul, Lai pö-kul, Shonshē ma-kul, Banjogi kūl, Shö [Chin] kūl, goi). There does not appear to be any trace of this word for a score in the Tibetan and Himalayan languages, where twenty is invariably rendered by “two-tens.”[4] The Northern Indian word kōṛī, which has the same meaning, has been compared with it; it is difficult, however, to imagine borrowing on one part or the other.

In the series of tens, 30 to 90, Mikir prefixes the multiplier: thòm-kèp, philī-kèp, phòngō-kèp, etc. The Boṛo group prefixes the tens (Garo sot-brī, 40, sot-bonggā, 50, sot-dok, 60, etc.).[5] The Naga group has both systems; Angāmi prefixes the tens: lhī[6]-dā, 40, lhī-pangu, 50, lhī-suru, 60; Lhota and Ao suffix them: Lhota tham-dro, 30, zü-ro, 40, rok-ro, 60; Ao semur’ 30, lir’ 40, rok-ur’ 60. In the Kuki-Chin group the majority of dialects prefix the tens (Thado and Lushei sōm-thūm 30, sōm-lī 40, etc.), and this is also the rule for Kachcha Naga, Kabui, and Khoirāo, as well as for all the languages classed by Dr. Grierson as Naga-Kuki. But the Shö or southern Chins not only have the same collocation as the Mikirs (thum-gip, mlī-gip, nghā-gip)—an arrangement which also obtains in Burmese,—but use the same words. This coincidence is very striking.

The word for a hundred, phārō, bears no resemblance to any word expressing this numeral in the Boṛo languages. It agrees with the Angāmi krā, Kezhāmā krī, Sopvoma krē, and in a remarkable way with the words used by the Southern Chins (Taungthā ta-yā = tarā, Chinbòk phyā = phrā, Yawdwin prā, Shö (Chin) krāt). It will be seen that phā- in Mikir, k- in the Naga languages, and ta-, ph-, p- and k- in the Chin dialects, are numeral prefixes, and that the essential element of the numeral is rā (Mikir rō) or rāt. It appears in this form, without any prefix, in several other Kuki-Chin languages.

Here should be mentioned a custom which obtains in Mikir of counting by fours; a group of four is chekē or chikē, which corresponds to the Boṛo zakhai (jakhai). This system is used for counting such things as eggs, betel-nuts, fowls, etc., of the same class; e.g. vo-tī chikē phòngō-rā ē-pum, 21 eggs (4 × 5 + 1): chikē phòngō-rā pum-thòm, 23 eggs (4 × 5 + 3). Possibly one language has borrowed from the other. (This method of counting by fours is common throughout the Aryan languages of Northern India, where a group of four is called (gaṇḍā.)

Our conclusion from these comparisons is that while Mikir has few coincidences, beyond those common to the whole Tibeto-Burman family, with the Boṛo group, it has many with the Naga and Kuki-Chin groups, and especially with the Shö or southernmost Chins.

Before leaving the numerals, something must be said of the prefixes which they exhibit throughout the Tibeto-Burman family. Taking first that member for which we have the oldest materials, Tibetan, the first ten numerals are as follows:—

As written.As now spoken in Central Tibet.
onegchigchik
twognyisnyī
threegsumsum
fourbzhishi
fivelngāngā
sixdrugḍhuk
sevenbdundün
eightbrgyadgyā
ninedgugu
tenbchuchu

Here we observe several different prefixes, once no doubt supplied with vowels, but from the dawn of written record united in Tibetan with the following consonant, and now no longer heard in utterance; in the first three units the prefix is g-: in four, seven, eight, and ten it is b-: in six and nine it is d-: and in five it is l-.

In the Tibeto-Burman languages of Assam and Burma we find the same phenomenon of numeral prefixes; but while some languages have the same prefix throughout the ten units, others, like Tibetan, have several different prefixes. In some cases, again, the prefixes have been incorporated in the numeral and are no longer recognized as separable, while in others they may be dropped when the numeral occurs in composition; in others, again, the prefixes have (as in spoken Tibetan) been dropped altogether.