8/ 



often very confusing species. Whether the anatomical structure of a plant, - - the number for 

 instance of long-celled rows in a periocl of intra-nodal grovvth, the height of the cells, and 

 the dropping off or persisting of the cortical layer in the node, - - offered characters of specific 

 value, was a question that had to be solved. The result was at fïrst not very satisfactory, 

 variations in the successive number of long-celled rows in a period of intra-nodal growth occur 

 so often, even in the sa me plant, that one must be cautious not to jurrip to rash conclusions, 

 and other characters already described at length by previous authors, for instance the size of 

 the conceptacles, the form of the froncl etc. must be taken in consideration before naming a 

 species. Still if the same constant structure is found regularly in different plants, this structure 

 may be a good help to recognize a species, for it is highly probable that such plants will 

 belong to the same species even when the outward form happens to be different. The node 

 of Amphiroa offered a character of importance and I will therefore describe it at full length. 



I began my studies on Amphiroa with specimens preserved in alcohol of A. rigida and 

 A. cryptarthrodia ■, these plants were authentic specimens of Solms-Laubach, kindly given to 

 me by Prof. Mayer at Naples. My methods of research were very simple ; the nodes, which 

 soon became a point of special interest to me, did not allow the use of ground sections; these 

 were only used now and than for comparison ; I decalcified my plants trying various liquids, 

 but for anatomical research Perenyi's fluid gave very good results and I used it continually. 

 The uncalcified material was stained, hardened, and often imbedded in celloidin. Other pieces 

 were sectionecl immediately as soon as they were hard enough ; the one thing I had to be 

 very careful of, was that my sections should go straight through the middle of the plant. If 

 they did not do this, I was not sure that the short cells of the branches, forming the cortical 

 layer, did not interfere with the succession of short and long cells such as really occur in the 

 middle of the plant. 



A. rigida and A. cryptarthrodia both have nodes consisting, as Solms-Laubach has 

 already observed, of two rows of cells. The node of A. rigida has been beautifully figured 

 by Yendo ; it consists of two cells which have the same size, and the ends of which in the 

 node rest not horizontally one above the other but obliquely. 



I thought this might tend to increase the power of resistance in the slender node. 



The node of A. cryptarthrodia consists also of two rows of cells, but here the lower 

 one is much larger than the upper one. In the joint the large cells have much more chalk in 

 their lower part than in the middle where it seemed to me that they were less calcified than 

 the cells of A. rigida. This has been indicated in the hg. 14, on PI. XVI, which was drawn 

 at the be^innino- of my research. In A. cryptarthrodia I observed also that the cortical layer 

 persisted much more regularly than I had supposed it would do according to Solms-Laubach. 

 These two al°ae had nodes so different from one another that a look at a section taken 

 through the node was sufficiënt to show to which species the specimen, from which the section 

 was taken, belonoed. This encouraged me. to pursue my research on the node and I submittecl 

 the whole Siboga material to an investigation on those lines. 



Amongst the collections of the Siboga were two species of which I had much material 

 at my disposal and naturally I began my research with these. My first impression had been 



