THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 179 



In Stud. 6 I quoted this species from Masatierra also. There were two 

 sheets of a Wahlenbergia in Herb. Lund; the}- had no labels, but on the sheets 

 was written in J. G. AGARDHs hand-writing: »Juan Fernandez, D. DOUGLAS 

 no. 48). As D. never visited Masafuera, 1 concluded that they came from 

 Masatierra. They did not belong to the species known before from Masatierra; 

 thus, only //'. Masafuerae remained, or a new species. Habitually they looked 

 like the former, but differed in the closer serrature of the leaves; that were 

 glabrous and had a somewhat different venation. The basal parts of the speci- 

 mens were missing. 



Comparing these specimens a second time with the rich material of W. 

 Masafuerae now at hand, I found that the differences quite forbid an identifica- 

 tion. I further found that, apart from these specimens, nothing like them had 

 ever been collected in Juan Fernandez, nor did such a species exist among 

 Douglas's plants in Herb. Kew. During my visit to that herbarium last year 

 I had a look at the Wahlenbergiae from Saint Helena, and was at once struck 

 by the great likeness between IV. linifolia A. DC. and the plants in Herb. 

 Lund. Some time ago I sent for both and was able to establish their identity. 

 The examination of a capsule of the alleged DOUGLAS plant revealed that it 

 is 2-valvous as in all the St. Helena species, while all Juan Fernandez species 

 have 3-valvous capsule. Certainly we can draw no other conclusion than that 

 the statement on the Lund specimens is quite false. And if we go back to 

 Douglas's journal, only lately published, we read under no. 48: »Campanula(r) 

 leaves alternate, sessile, lanceolate, serrate, smooth above, slightly pubescent 

 underneath » etc. Douglas no. 48 is W, Larraini. The person who distributed 

 the plants in question to Lund has made a rather dangerous mistake. 



Area of distribution: Endemic in Masafuera. 



115. W. Berteroi Hook, et Arn. -- JOHOW, Estud. 75. — Syn. Campa- 

 nula gracilis var. revoluta Colla 1. c. 119. — Plate 20, fig. 8. 



Masatierra: BERTERO no. 1440! »ad rupium fissuris in montibus editiori- 

 bus et in Goats island*. Rediscovered in Masatierra by us, but not in the 

 »higher mountains*: B. Cumberland, Pta San Carlos, two small specimens 

 (fl. 9 /i 17, no. 265); in the sand outside the caves, one large and beautiful plant 

 (fl. V12 16, no. 43). Both localities near the sea. 



Santa Clara: Bertero (»Goats island*); JOHOW! — Morro de los Alelies 

 (also observed by JOHOW), fr., some few fl. left, 26 /i 17, no. 345. 



IV. Berteroi has (just as W. Masafuerae) a large, lactiferous subterranean 

 bulb, formed by the base of the primary stem together with the primary root. 

 From the surface of this bulb spring the vegetative-floral shoots, each appar- 

 ently of one year's duration. The shape of the corolla is unlike that in other 

 species. Its colour is a deep rosa, the base crimson. Style and stamens are 

 rose-coloured, anthers golden yellow. The style is much longer than in other 

 species, the stigma visible in the mouth of the corolla. As the figure of 

 HOOKER and ArNOTT in Hooker's Journ. 1 (1834) plate 137 is incorrect ns 

 to the flower, another was prepared. The flower figured here was trimerous 

 throughout with 6 sepals etc., an anomality sometimes observed. 



Area of distribution: Endemic in Masatierra and Santa Clara. 



