104 



hampered his progress, but this statement should be received with 

 some caution. Of the 5 species named by him in illustration of it, 

 all of which are littoral plants, one is an erect shrub (Scaevola frutescens 

 Krause), two are herbs with creeping but no twining stems (Ipomoea 

 pes caprae Sw. ') and Canavalia rosea D.C.), the 2 others (Vigna 

 marina Merr. and Cassytha filiformis L.) are indeed twiners but both 

 are, like the 3 others practically restricted to the immediate neigh- 

 bourhood of the sea, consequently to the outer border of the grass- 

 jungle. None of these 5 species can at some altitude above sea-level 

 have occurred in considerable numbers, though some stray specimen 

 may have grown here or there. 



When P e n z i g and his companions explored the steeper parts 

 of the unweathered old basaltic slopes (from which the covering 

 layer had been washed away very soon after the eruption), the 

 aspect of the flora there appeared to be practically the same as 

 in 1886. In other words, just as might be expected, the same 

 kremnophytic vegetation was present that T r e u b had observed 1 1 

 years before -). Several species of ferns predominated in very many 

 gregariously growing specimens, numberless prothallia were seen, in 

 many places slimy blackish or bluish green Schizophyceae formed 

 still the only covering of the rocks and the soil. 



Of the components of the fern-association in 1897, P e n z i g, 

 in his description of the vegetation of the rocky localities, mentions 

 only 4, viz.: Ceropteris ca/ome/anos Und. (Gymnogramme calomelanos 

 Kaulf.), Acrostichum aureum L,, Stenochlaena pa/ustris Bedd. (A- 

 crostichum scandens J. Sm.) and Dryopteris setigera O. K. (Aspidium 

 vile Kze i ). From the collective enumeration of these ferns as plants 

 of the rocks clearly appears that Penzig did not pay to oecological 

 factors the attention required for a scientific investigation. As a 

 matter of fact these 4 species are never found together. Ceropteris 

 calomelanos (See p. 54 under No. 1 ) and Dryopteris setigera (See p. 57 

 under No. 8) are halophobous kremnophytes; they must have been 

 found on steep walls, in crevices of rocks or on ridges of pumice 

 and ashes. The two other ferns are inhabitants of swampy or humid 

 localities or at least of a soil with a high water-table. Acrostichum 

 aureum L. prefers a saline substratum 4 ) and occurs only exceptionnally 



1 ) Very rarely the stems of Ipomoea pes caprae show some tendency to twine, 

 but never they form a real obstacle. 



2 ) One should bear In mind that Treub took part in this second botanical trip 

 and could communicate his impressions to Penzig. 



:i ) See p. 58, sub No. 8. 

 4 j Cf. p. 55, No. 4. 



