TKOCHOrHORE OF HYDKOIDES UXCINATUS (EUrOMATUS). 555 



It appears to me, however, tliat in Phoronis, or at least 

 in the early stages of .the Actinotrocha larva, Ave have exactly 

 the same thing- as in Annelids. 



I have said that in Phoronis the mesoderm arises in the 

 region of the blastopore as a nnmber of irregular cells, which 

 are budded off into the blastocoel. These scatter throughout 

 the cavity, where they give rise to the mesodermic structures. 

 According to De Selys Longchamps (34), some of these cells 

 in the trunk region collect to form a rather imperfect coelomic 

 sac about the rectum or posterior portion of the stomach. I 

 Avas of opinion, however (36), that the cells that gave rise to 

 this sac had their origin in the gut wall, but of this I was by 

 no means certain. In any case, in Phoronis we have the 

 mesoderm showing a specialisation into a coelomic portion, 

 forming the primitive coelomic sac, and the irregular meso- 

 dermal cells scattered throughout the blastocoel. AVhether 

 we regard the coelomic portion as arising from the gut wall 

 or nor, it seems to me we have here the two forms of meso- 

 derm as in Annelids, and that Phoronis is intermediate 

 between Annelids and animals in which the mesoderm is 

 entirely ccelomic. Korschelt (22) sums up the mesoderm 

 formation under five heads, Avhicli are worth reviewing in 

 this connection. 



1st. Mesoderm band formation from teloblasts or pole-cells, 

 as in Annelids and Molluscs. 



2nd. Secondary mesoderm band formation, a modification 

 of the above process, and re-multiplied in Arthropods and 

 Cephalopods. 



3rd. Formation of mesoderm from gut pouches. 



4th. Formation of mesoderm from solid out-growths of the 

 gut. 



5th. The mesenchyme cells alone give rise to the coelom 

 and all the mesodermic structures. 



To the trochophore originally described by Hatschek (17) 

 in Eupomatus undue importance has perhaps been attached, 

 for such a trochophore is possessed by only a limited number 

 of Annelids, and is almost exclusively confined to the group 



VOL. 56, PART 3. NEW SERIES. 38 



