NOTES ON MISTLETOE 



295 



difficult to get a satisfactory photograph. I think it not improbable 

 that the seeds retain their positions on these threads, in the order 

 in which the berries were first swallowed. I have noticed that the 

 gluten in the berries w^hen in their natural state dries up very 

 quickly. If seeds are squeezed out and placed on a tree, by the 

 next day the gluten, which has stuck the seed down firmly, has 

 dried up and disappeared. The gluten on digested seeds remains 

 soft and sticky for two or three days (and it may be longer), and 

 the covering of skins is evidently intended to assist in preserving 

 this moisture until the rain comes and helps in the distribution. 



I have also proof that the Visciim seeds can lie dormant for 

 some time, just as other seeds often do. A seed placed on the 

 trunk of a copper-plum tree, at a spot where a small bough had 

 been removed, and where the bark had grown over the place, 

 leaving a mere dent in the centre, remained dormant for nearly 

 three years. The two " seed " leaves then appeared, and although 

 the plant looks quite healthy, its growth since then has been very 

 slow. The wood must be " end grain " where it is growing, and 

 the bark round it looks extremely hard. 



THE VEGETATION OF HARPTEEE COMBE. 

 By H. Stuart Thompson, F.L.S. 



On July loth, 1916, I visited Harptree Combe, N. Somerset ; 

 and was surprised to see so many fiowering plants and ferns 

 growing upon the stone retaining wall and massive pillars sup- 

 porting huge water-pipes belonging to the Bristol Water Works 

 Company, and built about seventy years ago. I made a somewhat 

 hasty list of about 77 species upon the stonework. 



The Combe, about a mile long, descends in a northerly direction 

 from slopes of the Mendip plateau tow^ards the beautiful Chew 

 valley. It is less bare and rocky and otherwise different from 

 most of the Mendip combes. Moreover a small stream, overgrown 

 with vegetation, flows along the bottom. It is wooded on both 

 sides, so that owing to the shade and moisture much Cystopteris 

 fragilis grows both on the stonework and under the neighbouring 

 trees and bushes, where the aqueduct crosses the valley. This 

 fern is very rarely seen growing on the ground in woods. It is a 

 less rigid form, as one would expect. Other ferns noticed on the 

 stonework were ScoIoiJendrmm, A. Trichomanes, A. Ruta-muraria 

 and one small plant of Bracken. There w^ere several mosses and 

 some Marchantia. The only grasses observed weie ArrJienatJieriim 

 aveiiaceum in quantity, Melica nutans L., Avena pratensis, Holcus 

 lanatus, Dactylis glomerata and Anthoxanthum odoratum. There 

 was a bush of Bhamnus cathai'ticus, one of Hazel, and a small 

 Beech and Ash. The most abundant large plant was Eupatorium 

 cannabimim, with a fair quantity of Campanula Trachelium and a 

 pale-leaved Hieracium not yet in flow^er. The list includes -^ra^e's 

 hirsuta, Cardamine hirsnj,a, Lychnis dioica, Cerastium viscosum, 



