846 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. XXIV. 



about one or two feet from the wall. When the room was opened on our 

 return ten weeks later, we found large portions of the outer surface of the 

 box covered with red earth, while a great deal of the wood had been eaten 

 away by white ants. The flooring was absolutely intact and there were no 

 signs of any communication tunnel to show where the white ants had come 

 from. There was^ however, a small quantity of red earth where the wall met 

 the floor about two feet from the box, so that the white ants must have 

 crossed this distance in the open carrying the very large quantity of earth 

 with which they plastered the box. Unfortunately they had ceased working 

 before our return. One's usual experience with white ants is that they are 

 averse to light and so from the point where they emerge from the ground 

 build a tunnel as a means of communication until they reach the object of 

 their ambition. 



N. C. MACLEOD. 

 Bombay, August, 1916. 



NO. XXIII.— NOTE ON AN UNDESCRIBED SPECIES OF CYNODON 

 BY K. EANGACHARI and C. TADULINGAM. 



{With a Plate.) 



While studying the grasses growing on the Farm of the Agricultural 

 College, Coimbatore, this grass attracted our notice as it dift'ered consider- 

 ably from the ordinary Cynodon dactylon, Pers. On examining the speci- 

 mens of grasses in the herbarium there were specimens matching this grass 

 that were collected by Dr. C. A. Barber in the Districts of Tinnevelly and 

 Godavari during 1901 and 1902, by M.Pt.Pty., C. Tadulingam in the Dis- 

 tricts of Kistna, Guntur and Nellore in 1907 and by M.R.Ry. Rai Bahadur 

 K. Rangachariar in the Chinglefjut District during 1898 and 1914. After 

 a careful study of this plant and after reference to all the literature on 

 grasses available in the Agricultural College Library, it appears to be to us, 

 an undescribed and unrecognised species. So we propose naming it 

 Cynodon barberi. 



Description. 

 Cynodon barberi, sp. nov. 



The plant is perennial. Stem are slender, radiately creeping close to 

 the ground 30 to 60 cm. long, rooting at the nodes, invariably with two or 

 three rarely more branches from each node ; flowering branches slender, 

 erect or ascending 2'5 cm. to 13 cm. high; internodes 2"5 to 6'5 cm., 

 slightly flattened, pale, green or purplish. 



Leaves — 1 to 3'5 cm. long and 3 to 4 mm. broad, flat, linear, acute or 

 subacute, scaberulous above and more so along the margins ; sheaths short, 

 smooth, compressed, with scattered long hairs at the mouth ; ligule a 

 narrow membrane with laciniate edge. 



Spikes — 3 to 5, slender, digitate, 2 to 4 cm. long, erect or spreading at 

 the end of thin peduncles, pale green or sometimes purplish ; rachis 

 slightly angular. 

 Spikelets — One flowered, compressed laterally, sessile or obscurely pedicel- 

 led, imbricate, alternately 2-seriate on the ventral side of the rachis ; 

 rachilla produced into a bristle behind the palea, with or without 

 minute pale membranous sterile glumes, a little longer than half the 

 spikelet and disarticulating above the involucral glumes. 

 Glumes three. — Glumes I and it (Involucral glumes) dinstinctly unequal 

 narrow, keeled with a strong keen nerve, lanceolate, keel shortly scabrid 

 Glume I 1'5 mm. long acute. Glume II 2 mm, acuminate equal to or 



