A XEW MARSH OHCHIS KjO 



We are tlierefore of tlie opinion that these plants all belong to 

 one type. VVe have seen Mr. (toss's plant at Kew, and judye that, 

 as far as may be determined from the dried plant, it is the same as 

 tliose from Arran and Hawkshead. Now in regard to this plant all 

 the considerations which make us doubt that O. piirpurella foi-m A 

 is to be referred to O. omenta apply, with this added that the form 

 of the lip is, for the most part, obscurely but detinitely trilobed. 



The first gathering which we had from Hawkshead was sent by 

 Miss V/ilson on June 17, 1916. It consisted of dwarf plants, 

 10-15 cm. high, with few, broadly-lanceolate leaves, in some cases 

 overtopping the spike, having the same very small spots as in the 

 case of the Aberystwyth plant, often only at the tips of the Is. A 

 niunber of plants sent from the same place on June 22, 1917, were 

 larger, 20-25 cm. high, but otherwise precisely the same: The flowers 

 a rich, dark purple, not the bright red-purple of the Aberystwyth 

 plant; lip slightly trilobed— this is more evident when the plant 

 is dry. The form varied from the rounded lip-type of O. prcetermissa 

 to the pointed type of form A. The leaf-pattern is of heavy crimson 

 marks. On July 8, 1916, we received from Arran a number of 

 plants, gathered by Mr. Allen, precisely similar to the Cumberland 

 ])lants. Tliey were growing in company with O. maculata and the 

 maroon form of O. incarnata v. dunensis. In both localities hybrids 

 with O. maculata occurred. 



With Mr. Goss, we should have called these forms a rather dwarf 

 variety of O. latifolia, had it not been that they were so evidently 

 to be considered along with the plants alread}' named O. crucnta. 

 At the same time the short habit and rather diamond-shaped lip, and 

 especially the peculiar type of leaf-spots, brought them into con- 

 nection with the Aberystwyth form. Finally, when we had considered 

 the difficulties in the way of ranking them as O. cruenta, we were 

 driven to grouping them together under a new specific name as 

 0. pmyvrella . 



It may be added that form A is more distinct from all others of 

 the group than form B, and we should rest our case mainly on that, 

 if 'Awj serious objection were maintained against form B. 



An average plant of form B, taken from the taller 1917 gathering, 

 is 26-5 cm. high, tubers included. Tubers narrow, tapering, divaricate. 

 Stem -sheaths lai-ge, dark at the tips, strongly veined. Leaves long- 

 lanceolate, blunt, lowermost small, second (broadest) 28 mm. broad, 

 third (longest) 1 dm. long, topmost bract-like, just short of the base 

 of the spike. Spike 3-5 cm., globose. Bracts broad below, strongly 

 veined, rather short. Flowers bright purple. Lip cuneate, about 

 8 mm. wide by 6 mm. long, crenulate below, centre-lobe blunt, small, 

 markings heavy, of dark crimson. Sepals erect, spur of medium 

 length, slightly tapering, modeiately stout. 



HyhricU of O. pnrpurella. We have so determined some very 

 beautiful plants growing in fair numbers with form A. They 

 appear to be crossed with a dwarf form of O. latifolia, which is also- 

 growing in the vicinity. A few ])lants which appear to be crossed 

 with O. erief'toriiDi also occm'. 



(Jf form I), both fi'om Arran and llawksliead. we have received 



