28Ö 



found in each conceptacle, but Iheie must be some reason why it 

 (loes not eslablish ilself afler hairs have been formed. No trace 

 of a development of Ihe »initial» was found; also in the earliest 

 siages (lig. 4) it is niature and shows Ihe long free lail. The posi- 

 tion of Ihe epiphyto will naturally be central; when, in Ihe growing 



Fig. 5. Section of a somewhat oider oonceptncle oul}' lateral part 

 shown), X 3(J0. 



apex of a branch, we find half a dozen of these bodies, I believe 

 most of them have found Ihe weak spöt where a conceptacle is 

 soins to be formed. 



Conclusions. 



Kjellman (Nat. Pflanzenfam.) includes Splachnidiiim in Ihe Fiica- 

 ceae, without much discussion. Mitchell and Whittixg have an 

 order Splachnidiaceae, intermediate, they say, between Laminariaceae 

 and Fiicaceae. Roi-: retains this position: »as to the vegelalive 

 growth, the tissue bears an ecjual resemblance to the Fucaceae and 

 Laminariaceae, the only essenlial difference being in origin.» She 

 judges Ihe conceptacles to be a good fucaceous character; still 

 »there are deep furrows in Macrocijstis and Postelsia, which closely 



