1890.] NEW YORK ACADEMY OF SCIENCES. 17 



the centre, Avhicb in others was filled with calcareous material. 

 I believe Dr. DoUey's theory of the formation of these cylin- 

 drical and tubular projections is, in the main, the correct one; 

 for I collected specimens with the roots still in them, and the 

 concentric arrangement shown on cross section shows very 

 plainly that the solvent action of the water, following the path 

 of the root, had been from within outward. As these masses are 

 conspicuous and characteristic, I would suggest that they be 

 called rhizomorphs ; and this name would also cover the tubular 

 masses that have evidently been formed in the same manner, 

 and which I have found in the ferruginous clays of New Jer- 

 sey and elsewhere. The action of the water in the latter case 

 has been on the iron contained in the clay. 



Dr. Dolley is inclined to think that the rhizomorphs were 

 formed while the sand was in a loose state; and while I do not 

 deny that this might be possible, I believe that all that I saw 

 had been formed by the roots penetrating the already hardened 

 rock; and I furthermore found no rhizomorphs being formed 

 by the plants growing in the loose sand. 



Dr. Dolley also states ' that the "small islands exhibiting these 

 peculiar formations " (rhizomor))hs) "are indications, therefore, 

 of erosion and subsidence.'' I admit the erosion; but ap the 

 rhiz(miorplis are found twenty and thirty feet above the sea-level, 

 I think that subsidence has taken no part in their formation. 



These rhizomorphs are not to be confounded with the cylindri- 

 cal masses described on Goat Cay, for the latter were vertical 

 and did not exhibit the concentric arrangement of the particlea 

 so plainly shown in the rhizomorphs, which are found at all 

 possible angles. And if we suppose the masses on Goat Cay to 

 be due to the same i)rocess that formed the rhizomorphs, we must 

 suppose that the roots that produced them grew vertically down- 

 ward, which is extremely improbable. It may be well to state 

 that a few rhizomorphs were found on Goat Cay, and the differ- 

 ence between them and the cylindrical masses was very marked. 



It might be asked, if these rhizomorphs have been formed 

 in the manner described, why is it that we do not find them 

 everywhere on the islands ? And in answer to this objection it 

 may be said that roots of most of the trees spread out over the 

 ground or slightly under the crust, but do not penetrate the 

 rock; and that the roots of the smaller plants are those that 

 have formed the rhizomorphs, and that these latter show only 

 where tiiere is but little vegetation, and where the erosion is 

 active and not helped by tiie roots of trees breaking up the sur- 

 face into blocks as described above. 



'Loc. cit., p. 132. 



