78 Lloyd and Tracy: The insular Flora 



tropical strand succulents have been described by Warming. The 

 salicornias, which are plants of more northern distribution, are by 

 no means so plentiful as Batis. They are, however, of the same 

 ecological character, being stem succulents, while Batis is a leaf 

 succulent. The only other plant occurring in the muck marshes 

 is Aviccnnia nitida, with its well-known negatively geotropic roots 

 (pneumathodes). This mangrove grows abundantly on some of 

 the islands to the extent almost of the exclusion of other species 

 (Breton Island), but in no case reaches a height greater than 60 

 cm. They produce, however, flowers and fruit. On some of the 

 islands visited great numbers of dead plants were found which 

 had doubtless been killed by the frost. The fact that all the dead 

 ones were about of a size indicates that this plant is so killed and 

 that it is here in its northern limit. The soil of the salt marsh is 

 composed of a very finely-divided alluvium which packs very 

 tightly, and this with the water which is saline results in a mass 

 which effectually prevents the entrance of air. This accounts for 

 the development of the pneumathodes of Avicennia and the abun- 

 dant aerenchyma of Batis, paralleling the well-known structures 

 of similar adaptive significance in many swamp plants. 



The general appearance of the salt muck marsh vegetation is 

 one of somber uniformity. The succulents are of the same height 

 and color — a dark green, and the grass association is a bluish gray. 

 The variety of color of the northern salt marsh plant covering is 

 well-nigh absent. 



From the above it will be seen that there are scarcely any plants 

 common to the shell strand and muck marsh. Apparent excep- 

 tions (e. g. t Spartina juncformis) may usually be explained by the 

 fact that the roots may penetrate into the mud through the 

 shells. Marsh plants may thus be found in the shell strand, espe- 

 cially in the shell-muck marsh. The converse is scarcely ever 

 true. The lee edge of the encroaching shell dune shows a sudden 

 and complete change from the muck marsh, as is shown in 

 Plate 4. 



Shell Marsh Formation 

 In some instances the shell dune becomes broad and scattered 

 on the inner limit and the shell fragments are mixed with the allu- 



