38 



of wliich two more tubers arc sliowii on plate ii, at the top on the 

 right. 'Ilicse tubers were usually deeply fingered, which is not 

 such .IS the race had i)roduccd in IDIG (vide Plate ii, of the Bul- 

 letin. \'(il. 1, (h)uhle .\'o. 11-12). It may be that twinning and 

 grouping of tubers is the result of an arrest of the first attempt at 

 tuber-formation in such a case; but this is a guess only to be 

 investigated by cvpcrimentally varying the conditions in plant- 

 ing. '1 he upper block of ])late ii. illustrates api)roximately the 

 condition which Sir David Prain and the writer named " fareini- 

 formis " or sausage shaped, — a condition common in the Gangetic 

 valley and in other parts of India, which has not been figured in 

 the previously i^rinted ])lates. It is seen that when the race X^. 

 ■((i, is at one of its extremes, namely that on the right of the Plate 

 ii, it is such as can reasonably be classed as farciniformis but at an- 

 other extreme it is far away. This observation is a caution against 

 accepting il'l-(U'Vcl()pcd tubers as ])roving their race to belong to 

 " farciniformis." 



h*ace 76 lias other characters than those of shape which must 

 be remarked: its flesh is peculiarly firm and of the yellow of old 

 ivory. Cooked it was i)ronounced fair. It is a heavy yielder, with 

 a prickly stem. 



Continuing the comments on the plates, and turning to the 

 middle block of plate ii, four tubers of race 132, — the Fijian " Uvi 

 kaboa " — are seen, above four tubers of the race No. 44, which was 

 received from the Bureau of Agriculture, Philippine islands. These 

 illustrations will ser\e as references, so that it may l)e possible in 

 descriptions lo point to them saying, " fingered to the degree of Xo. 

 i;)2," or " fingered to tlie degree of Xo. 44." 



'J he hottoui block of pbite ii. has a historical l)earing, iov it 

 exiiil)its what Humph i us calletl i'bi tangan or Hand yam, and L'bi 

 iilar or Snake yam, the first on tlie left in races 14 and T4, and tne 

 second on the right in race o7<). All tihree came from the Philip- 

 |)ine islands. 'I'o he really like a liand, the tubers must be well grown ; 

 fasciation wberefrom the flatness conies not developing unless the 

 tuber is strong enough to branch freely. Therefore the tubers on 

 the extreme left which are not well develojied, do not show it. 

 b'uniphius hitherto has been much misunderstood in regard to L'bi 

 ubir: li'oxbuigh though it must be a wild yam other than Dioscorea 

 alala, and called a most unlike Indian s])ecies " Dioscorea anguina " 

 in consetpience (Flora indica. l<So2, iii. ]i. Soii). 



It is (piile cci'tain that Ihi ular is Dioscorea Khihi. and as that 

 tuber shows which in the plate is in contact with the half-metre 

 measure, it should be classed along with those ra<-es which do not 

 bury their tubers by desceiuling into the earth, but may extrude 

 them, — tho.se which have to be cultivated i)y earthing-u]). 



J'lates iii, and iv, are hath given to illustrate this peculiar 

 group, — the not-burying or apogeotropic grou)). Plate iv, illus- 

 trates it at its extreme I i)late iii, shows races which are inter- 

 niediate between the more usual state and it. Firstly on the upper 



