(Fawcett 1913).

"Plants small, the pseudostems 1.0 - 1.5 m. tall, suckering freely ; petioles 5.0 - 8.0 cm. long, grooved with incurved scarious margins, the basal part abruptly expanded into large corrugate black auricles, the auricles 10 - 20 cm. long, 3-4 cm wide ; leaf- blade 65 - 150 cm. long, 20 - 30 cm. wide, oblong lanceolate, bright green, apex acute or truncately obtuse, base obtuse and shortly decurrent to the petiole, the two sides at the base equal or slightly unequal ; inflorescence erect ; peduncle and rachis densely pubescent ; sterile bracts usually 1, deciduous ; 5 - 6 basal hands female, upper hands male ; male bud in advanced stage of blooming ovoid and top-shaped, the bracts imbricate ; bract purplish pink, shining, oblong-rhombic, attenuately narrowing into the tip, boat-shaped as a whole, thick and coriaceous in the centre, thinner near the edges, slightly deflexed at the flowering time, not recurved backwards, soon deciduous ; male flowers persistent or rot and drop off from the rachis in wet condition, 8 - 12 per bract in two rows, the abortive ovary 3 mm long, the compound tepal 3.5 - 4.5 cm. long, ribbed at the dorsal angles, with 5-toothed apex, the outer lobes ovate, 3 mm. long, with hairy filiform appendage 3 mm. long, the inner three lobes thin, glabrous, acute, the central lobe larger than the laterals, 2 mm. long, the free tepal oblanceolate with bittenly [sic] obcordate apex, transversely corrugated at the basal part, 3.0 - 3.5 cm. long, 1 cm. wide at the widest part near the apex, the stamens 5, nearly as long as the compound tepal, filament 2.6 - 3.0 cm. long, fruit bunch small, erect and loosely packed, consisting of 5 - 6 hands of 3 - 6 fruits each ; fruits connated laterally about two thirds of its length, densely covered by the soft yellow hairs, individual fruit bottle-shaped, ridged, 4.5 cm. long, 2.5 cm. in diameter ; seeds cylindrical obpyriform, 3 mm. in diameter, 5 mm. long, longitudinally and obscurely ribbed with many tubercles".

(Hotta 1967).

Simmonds 1960 described M. hirta as "a small plant with erect bunches, red bracts, hairy coherent fruits and angular sub-globose seeds" and commented that though in aspect it "must resemble a Callimusa the seeds disagree and its affinity may turn out to lie with Musa beccarii".   This discrepancy is explained by Hotta (1967):

"By the shape of connated fruit with dense yellow hairs, the small size of plants, number and arrangement of the flowers, our specimens agree very well with Beccari's description of M. hirta. There is, however, discrepancy in the seeds between the original description and our specimens. Beccari described as "semina parva, 4 mill. longa numerosa, late ovata vel irregulariter globosa vel obtusissime angulosa, rugoso-tuberculata." Judging from this description, this species has no position in the Sect. Callimusa as mentioned by Simmonds, but our materials have cylindrical obpyriform seeds, which are undoubtedly the feature of the Sect. Callimusa. By the very unique and distinct characteristics of fruit, our plants are referred with no doubt to M. hirta. Aforesaid discrepancy may presumably lie in the variability in this feature".  It would be very odd for seed shape to be so variable.

The sectional status of Musa hirta is currently problematic.  If its affinities really are with M. beccarii this rather suggests it belongs wherever that rather better known species belongs.  Simmonds and Weatherup 1990 transferred M. beccarii to section Callimusa out of incertae sedis.  Jong & Argent 2001 maintained M. beccarii as incertae sedis.  Based on its seed structure, chromosome number and AFLP results, Wong et al 2001 "confidently placed" M. beccarii within section Australimusa.

So is M. hirta also section Australimusa?  The discrepancy in seed shape between Beccari's material and Hotta's is worrying but Hotta's description is the best available and taking it at face value it seems reasonable, for the time being, to leave M. hirta in section Callimusa.  Although they did not look at M. hirta it may be relevant here that Wong et al suggest that section Australimusa and Callimusa can no longer be maintained as distinct.

Musa hirta is one of only six bananas currently known in which the fruit splits (or dehisces or is schizocarpic) on maturity, the others are Musa velutina from India, Musa johnsii from Papua (Irian Jaya), Musa lolodensis and Musa schizocarpa from Papua New Guinea, and Musella lasiocarpa from China and northern Indo-China.