109 



It seems to me that just as closely as Amphiroa is allied to Lithothamnion, so closely is 

 one of the new genera allied to Goniolitkon and I called it therefore Metagoniolithon. In both 

 we find cells of a uniform size, in Goniolitkon throughout the whole frond, in Metagoniolithon 

 with this difference, that the cells in the node are smaller than in the joint ; in both, the cells 

 communicate by rather large thin places in their longitudinal walls and in Goniolitkon frutescens 

 I observed slight constrictions of the frond, that seemed to indicate how easily nature might 

 develop this constriction into the node of Metagoniolithon. 



In Goniolitkon however I observed Deckzellen, not quite as such a regular layer as in 

 Lithothamnion or Amphiroa, but still each apical cell had its own "Deckzelle". In young fronds 

 of Metagoniolithon graniferum I did not see any "Deckzellen". 



It struck me as very interesting to observe how from two genera so closely allied as 

 Lithothamnion and Goniolitkon had probably sprung forth two genera Amphiroa and Meta- 

 goniolithon which resembled one another outwardly so much and were still so different. 



Fossil forms. 



The study of the anatomical structure of the Corallineae verae afforded the means of 

 recognizing at once in sections of fossil calcareous algae, whether the plants seen in the sections 

 belonged to Amphiroa or to either of the other genera. The tissue of the nodes, not being 

 calcified, has always disappeared but the nodes remain visible in the sections as an interruption 

 of the calcareous tissue. 



In sections, that I owed to the kindness of Prof. Martin at Leiden, I found two plants 

 figured on pi. XVI, fig. 17 and 18. The sections were made from specimens collected on the 

 mountain Nona in Ambon and at Totok in Celebes. The section of Ambon bore the number 327 

 of Prof. Martin's collection; the section from Totok the number 44 coll. F"orster. Prof. Martin 

 was so kind as to teil me that the stratum at Ambon from which the section N° 327 came, is 

 situated 400 m. above the level of the sea and belongs probably to the pliocene. The stratum 

 at Totok in Celebes, section N" 44 belongs undoubtedly to the tertiary formation and is certainly 

 not younger than old miocene.. 



The plant from Totok shows us two half joints and an open space between these two, 

 where most probably the node was to be found when the plant was in a living state. When 

 we compare fig. 18, pi. XVI with fig. 15 on the same plate representing a recent Jania, 

 every one can see that the plant from Totok is more like a Jania than a Corallina. 



The length of the node, which is almost the same as the length of the cells in the 

 joint, is a character that points to Jania, for the only case that I know of Amphiroa having 

 nodes consisting of one row of cells, is A. valonioides and here the layers of cells succeeding 

 one another grow smaller and alternate as they should do in an Amphiroa ; but this they do 

 not do in the plant from Totok. 



To which species of Jania the plant from Totok stands nearest, I should not dare to 

 decide for the species of Jania are very difficult to know from one another, even in a living 

 state. The fact that the subgenus Jania occurred already in old miocene, appeared interesting 



