iiulgaaende undersogelser er i den senere tid toretaget af 

 Joliei (6) og Seelif/er (2 og 3), sum begge l)eskriver jieri- 

 hninrhialcaviteter og nervesystem som inesodermale dannelser. 



Knopperne hos Pyrosonia anlregges, som det a f de 

 nævnte arbeider er bekjendt, ved at moderdyrets eitiodiTiit 

 Inigter sig ud ved endostylens nedre ende, idet den danner 

 et tyndt ror, som gjennem eleol)lasten strækker sig helt 

 lieninKid eJdndrrnieii, der nu ogsaa begynder at bugte sig 

 ud. Entodermroret er noget sanimentrykt i dorso-ventral 

 retning, og det viser sig paa snit at have en rektangulær 

 form; ektodermen bestaar i knopskydningsregionen af ku- 

 l)iske eeller med jevn overgang til pladeepithelet. som ellers 

 omgiver moderdyret (Se fig. 5). Mellem disse lag tindes 

 paa undersiden af knopanlægget en temmelig kompakt me- 

 sodermceVemasse, hvori enkelte celler, de senere ægceller, 

 udmerker sig fremfor de øvrige ved sin størrelse. Af me- 

 sodermelenienter tindes endvidere. efter Seeliger, to mesen- 

 r//(/)i(-strenge, som strækker sig langs entodermroret paa 

 Ijegge sider, og desuden enkelte spredte celler. 



Ferihranclnalcaviteteriie optræder allerede paa et sta- 

 dium, hvor knoppen kun viser sig som en svag udbugtning 

 af moderdyret, og de stammer, efter Sedigerx opfatning. 

 afgjort fra niesoderuini. hvorimod lian ikke udtaler sig be- 

 stemt om. enten de skriver sig fra de meseneliymstrenge, 

 som ligger langs entodermroret, eller fra den større celle- 

 masse [laa knopjiens underside. 



Nervesystemet anlægges, efter SeeJigers undersogelse, 

 paa et noget senere stadium, ved at mesodermcelJemassen 

 strækker sig forover i knoppen, rundt entodermrørets spids, 

 saa at en del af den kommer til at ligge paa knoppens 

 overside. Denne del, den neurale mesoderm, afsnores snart 

 fra den underste, den lilimale, og samtidig antager den 

 formen af et ror, nerrerøret, idet der dannes en spalte i 

 dens indre. 



Angaaeude denne forste oprindelse til p^ribrancJtiaJ- 

 caviteter og nerverør er jeg kommen til andre resultater 

 end Seeliger, og jeg skal, under min behandling af de for- 

 skjellige stadier af knopper, specielt tåge hensyn til disse 

 organer, idet jeg, hvad knoppens øvrige udvikling angaar, 

 henviser til SeeJigers indgaaende beskrivelse. 



Stadium I. 



Jeg begynder med at beskrive et stadium som det 

 paa tig. 7 afbildede, hvor knoppen viser sig som en meget 

 tydelig udbugtning af moderdyret, men hvor nogen afsnoring 

 endnu ikke er begyndt. Eldodernien bestaar lier, som paa 

 senere stadier, af kubiske celler, medens entodermroret 

 begrænses af cjlinderepithel. 



Eidodermrøret er firkantet, men med indboiede side- 

 Hader, saa det paa snit viser sig korsformet. Paa denne 

 niaade dannes mellem ektoderm og entoderm fire parallelle 



Vun inrske Nordlnivsexpeditiin. Kristine Bonnevie: Dist.iplia rna; 



tiiorough investigatlons have lately becn made by Joliet (6) 

 and Seeliger (2 and 3), botli of whom describe the peri- 

 hrnncJiicd cavities and tiie nerrons sgsfem as mesodermal 

 forma tions. 



As is vvell known from the above-named works, the 

 buds in Fyrosoina originate in the evagination of the eH- 

 doderin of the mother animal at the iower end of the en- 

 dostyle, thus forming a narrow tube wiiieh e.vtends through 

 the elæoblast up to the crfoderm, wiiich now also begins 

 to ])ulge. The endodermal tube is somewiiat compressed 

 in a dorso-ventral direction, and exiiibits in section a rec- 

 tangular shape. In the budding region, the ectoderm consists 

 of cubical cells with an even transition to Hathened epithe- 

 lium, which also envelopes the mother animal (See fig. 5). 

 Between these layers, on the Iower side of the incipiont 

 bud, there is a fairly eompact mass of mesoderm ivds. in 

 which certain cells, the future egg-cells, are distinguished 

 from tlie others by their size. According to Seeliger, there 

 are, moreover, of mesoderm elements, two mesenchymal 

 cords, extending on both sides along the endodermal tube 

 and a few scattered cells. 



The peribranchial cavities have already made their 

 appearance at a stage when the bud only appears like a 

 slight sweUing of the mother animal, and, according to 

 Seeliger, they undouljtedly originate in the mesoderm. thougli 

 that writer gives no decided opinion as to whether they 

 are from the mesenchymal cords, or from the large cellular 

 mass on tlie under side of tiie l)ud. 



According to Seeliger, the nervous sgstem commences 

 to form at a somewhat later stage, by the expansion of 

 the mass of mesoderm cells forwards in the l)ud, and round 

 the point of the endodermal tube, so that a portion of it 

 reaches tiie upper side of tiie bud. This part, the neural 

 mesoderm, is soon cut off from the Iower, the liæmal meso- 

 derm, and simultaneonsly assumes the form of a tulie — 

 tile nerve tuhe — , a cleft being formed within it. 



With regard to the origin of tiie perihrancJiial cavities 

 and the nervetidie, I have arrived at a different conclusion 

 to Seeliger. and shall, in treating of the different stages of 

 huds, pay special attentiou to these organs, while as re- 

 gards the development of the rest of the bad, I would refer 

 the reader to Seeliger^s detailed description. 



Stage I. 



I will Ijegin by descriinng a stage like that represeii- 

 ted in fig. 7, where the bud appears as a very distinct 

 evagination ot tiie mother animal, liut where no separation 

 has as yet begun to take place. Tlie ectoderm, here as in 

 later stages, consists of cubical cells, while the endodermal 

 tul>e is l)ounded liy cylindrical epitiielium. 



Tiie cndddermal tube is quadrangular, but with in- 

 curved lateral surfaces, so tiiat, in section. it has the sliape 

 of a cross. In tiiis wav. there are formed between tiie 



gnilarva og Pyrosoma elegans. 



