24-2 THE JOUHNAL OF HOT AN V 



sheaths at the base of the stem were spotted or streaked with red. 

 (2 j The mid-lobe of the lip was often so much reflexed as to be 

 parallel with the under surface of the flower. (3) The base of the 

 lip and the centre of the mid-lobe were densely covered with erect 

 red hairs. (4) The whole lip was concolorous — a dark livid red, not 

 yellowish in the middle, and, more important than all the above, 

 (5) There were 2 parallel calli at the base of the lip, not one single 

 callus as in lingua. As 1, 2, and 5 are characters of 8. parviflora 

 Pari, (occultata Gray), I at first thought the plants must be hybrids 

 between it and S. lingua. Later on the same plant sprang up in 

 thousands, nearly always in colonies. It was no longer possible to 

 regard it as a hybrid, not only on account of its great abundance and 

 uniformity, but also because not a single specimen of either lingua 

 or parviflora could be found in the locality. It is true that both 

 these species existed in very limited numbers in moist sandy ground 

 near the sea, but their stations were separated from those of greaaria 

 by a wide plain several kilometres across, and moreover greaaria did 

 not occur in their neighbourhood. 



On April 30th I took examples of the Hyeres plant, now rapidly 

 going out of flower, to Vence, where on May 4th I found true 

 lingua in early flower. It was easy to see that gregaria differed from 

 it even in general appearance. The lip in lingua was always attractively 

 coloured, bright rose, yellow, or dark vinous red, usually yellow 

 between the side-lobes, and directed forwards and down wards," never 

 reflexed backwards. The flower was more open, the side-lobes pro- 

 truding from the helmet, and the callus easily visible. In qreqaria 

 the lip was always dark red, often so reflexed as to be hidden' beneath 

 the flower, and the callus could not be seen without pulling the 

 flower to pieces. 



I was therefore forced to the unexpected conclusion that there 

 existed in France a hitherto-unrecorded species of Serapias, and 

 decided to name it Serapias gregaria, from its habit of growing in 

 colonies. This is due to the annual production of a third tuber, so 

 that the plant is succeeded, accidents apart, by two new plants 

 instead of by one only, as in the majority of orchids, resulting in a 

 rapid vegetative increase. Lingua occasionally has 3 bulbs, in gregaria 

 it is the rule. It should be noted that gregaria flourishes upon the 

 schist, in company with Orchis Ohampagneuxii and O. saccata, 

 which are not found off that formation, whilst, at any rate at 

 Hyeres, lingua and parviflora occur, though only sparingly, on moist 

 sandy ground near the sea, where Barlia longibractcata, Ophrys 

 apifera, and O. arachnitiformis are rather abundant, all which 

 plants are absent from the schist, though frequent enough on the 

 limestone. It is difficult. to believe that lingua and parviflora could 

 have given rise to a hybrid which finds its' home on an entirely dif- 

 ferent geological formation. 



Plate 560, representing (A) 8. gregaria and (13) 8. lingua, shows 

 the differences between them better than a verbal description. The 

 following points should be noted in respect of A:- — (1) The red upper 

 stem. (2) The linear leaves tapering to a very fine point. (3) The 

 longer, more acuminate, and darker bracts. (4) The smaller, less open 



