WITHIN TUK GENUS BENDUONJBPHTHVA. 



43 



D. 2)ectmata must be regarded as the climax in the line of 

 evolution. 



Again in the rubra group, grade I. is well represented by 

 D. repens, while in the spinulosa group, grade II. is probably 

 represented by D. spinulosa itself with 6-8 pairs very small and 

 not projecting, even better by this than by D. flahellifera where 

 we have 8-9, but in which I recognize what I regard as the 

 beginning of differentiation : viz., a specialization of the highest 

 pair. 



In tabular form my conclusions may be represented as 

 follows : — 



II. 



Grade. 



VI.^ 



IV.- 



Ill, 



11.^ 



GlOMERATES. 



DiVAEICATES. 



III. 



Umbeliates. 



D. punicea (Stud.) *. f 

 D. carnea(Wr.&Stud.:). | 

 D. doederleiiii (Kiik.). [ 



D. piitteri (Kiik.). 



I), stiensoni (Holm). 



D. orientalis (Heiid.). 



D.flammea (sp. u.). 



J>. ce>'yicorw?s( Wr. & Stud.). | 



L>. laxa (Wr. & Stud.). L_ 



f; D. lutea (Kiik.). 

 I I D. longicaiilis (Kiik.). 

 J i I), simplex (sp. u.). 



D. spiuit'era (Holm). 

 D. kollikeri (Kiik.). 



D. studeri (Ridley) 

 D. gigantea (Verr.' 



D. mayi (Kiik.;. f 



D. liemprichi (Klunz.). \ 



I), argentea (Kiik.). 

 D. clavata (Kiik.). 



D. eburnea (Kiik.). 



D. mollis (Holm). 

 D. cirsium (Kiik.). 

 D. iuvoluta (Kiik.). 

 D. mirahilis (Hend.) 



D. Jclunzingeri (Stud.). 

 -D. ehrenhergi (Kiik.). 

 D. japonica (Kiik.). 

 D. microspiculata (Piit.). 

 I), eriiiacea (Kiik.). 



B. coj'OMaf«(Wi'.&Stud. 



D. disoifoi'mis (Kiik.). 

 D. ruln-a (May). 

 D. pumilio (Studer) . 

 D. uigrescens (Kuk.). 

 D. collaris (Wr. & Stud. 

 D. sinensis (Piit.). 



D. dendropliyta (Stud.). 

 D. hyalina (Kiik.). 

 D. mexicana (Kiik.). 

 D. tiorida (Esper). 



D. brevirama (Burcli.). 

 J), umbellulifeia (Kiik.). 

 T>. spinulosa (Graj-). 

 D. hahereri (Kiik.). 

 jD. annectens (sp. n.). , 



D. villosa (Kiik.). 

 D. australis (Kiik.). 

 D. repens (Kiik.). 



* The species printed in italics are those worked out from the DendronepMhya collection 

 of the ' Siboga ' Expedition. 



As a convenient means of expressing the composition of the 

 anthocodial point, I have drawn up an anthocodial formula on 

 the following lines : — The spicules of the point are denoted by 

 " P " if big and strong, and by " p " if small and weak ; those 

 of the ci'own by " Or '', and the supporting bundle by the letters 

 " S.B." preceded by a qualifying adjective, such as weak, medium, 

 very strong, etc. In this way the anthocodial grade and formula 

 of D. annectens, for example, can be set forth briefly as : 



II. = 8p + 0Cr + very weak S.B. 



