46 



(lets sitiis, anlæggfs noget for-kjellig hos hver familie; saa- 

 ledes dannedes det hos DidctpJia som en median udbugt- 

 ning af den indre blæres dorsalparti. hos AuiarouciuiH 

 anlagdes det sammen med venstre periljranciiialsækanlæg, 

 hos BotryUus sammen med det hele sadelformige peribran- 

 chialsækaulæg {Pyrnsoma, se den specielle beskrivelse). 

 Disse forskjellige mindre afvigelser er dog af liden betyd- 

 ning. Overalt er det vresentlige 1) at centralnerves3'stemet 

 overhovedet anlfegges som et ror, 2) at det indeholder an- 

 lægget for det blivende ganglion og hypopliysis (men ikke 

 for nogen larvehjerne), 3) at det fremgaar som en udbugt- 

 ning af den indre blære. Denne udbugtning vokser overalt 

 bagenifra forover først som en blindt endende cylinder, 

 vokser saa sammen med gjælletarmen fortil, afsnøres igjen 

 bagtil og synes saa, idet det snart komnuinicerer med gjælle- 

 tarmen, at danne en bagover rettet udbugtning af denne. 

 Ligesom hos larverne afsnores gangliet fra en af det ror- 

 formige anlægs vægge, og saasnart dette er skeet er dermed 

 ogsaa roret omdannet til hypophysis. 



Forskjellighederne i de to udviklingsmodi viser sig 

 derfor væsentlig at beståa i følgende: 



1) Udviklingen er mere compliceret f,.længere'') i larve- 

 udviklingeii end i knopudvikliiigen. 



2) Anlægget tager sin begyndelse fra et andet sted, en 

 anden del i knopanlægget end i det embryonalo anlæg. 



Saa store end afvigelserne nu er hos knopi)er og lar- 

 ver saa tindes der dog ogsaa lier lighedspunkter. Jeg 

 fremhrever saaledes : 



1) Anlægget til centralnervesystemet er baade i knop 

 og larve rorformigt iraedullarrøret, dorsalroret). 2) Dette 

 anlæg differentierer sig vistnok paa meget enklere maade 

 hos knopperne end hos larverne, idet ikke larvelijernen 

 overhovedet kommer til dannelse, men de dele af central- 

 nervesystemets organer, der er de blivende iios det udvok- 

 sede dyr, nemlig det blivende ganglion og hypophysis, de 

 differentierer sig ud fra det rørformige anlæg paa samme 

 maade i knop- og larveudviklingen. 3) Det rorformige an- 

 lægs forbindelse med gjælletarmen fflimmergruben) opstaar 

 baade i knop og larve secundært ved samnienvoksning. 



Aldeles de samme love tinder vi \m iov periliranchiaJ- 

 sækkens dannelse. 



Hos larverne saa vi, at peribranchialsækken anlagdes 

 af 2 dorsale indbugtninger af ektodermen. Med sikkerhed 

 tVemgik dette af litterat nren og egne undersogelser for de 

 fleste grupper og Pitoiis afvigende resultater for enkelte 

 andre gruppers vedkommende fandt vi iallefald forelobig 

 uantageligt af theoretiske grunde. Hos knopperne forlmldt 

 det sig imidlertid paa den maade, at peribranchialsækken 

 overalt dannedes ved udbugtning af den indre blære, hos 



which we here called by the neutral name. ..dorsal tube", 

 on account of its position, commences somewhat differently 

 in eacli family. Thus in Distaplia, it is formod as a me- 

 dian evagination of the dorsal part of the inner vesicle; 

 in Amaroucinm, it commences together with the left rudi- 

 ment of the peribranchial cavity; in BotryUus, together 

 with the whole saddle-shaped rudiment of the peribranchial 

 cavity. (For Pyrosoma, see the special description). These 

 various smaller deviations are, however, of little importance. 

 The most important facts in all cases are (1) tliat the 

 central organs of the nervous system commenee as a tube, 

 (2) that they contain the rudiment of the permanent gan- 

 glion and hypophysis (but not of any larval brain), (3) 

 that tiiey are formed as an evagination of the inner vesi- 

 cle. This evagination always grows from behind forwards, 

 at first as a blind-ending cylinder; it then coalesces witli 

 tlie branchial gut in front, is again constricted beliind. and 

 tlieu seenis, as it soon forms a communication with the 

 branchial gut, to form a backward-directed evagination of 

 the latter. As in the larva, the ganglion is constricted 

 from one of the walls of the tubular rudiment, and wheir 

 this has tåken place, the tube is thereby converted into 

 the hypophysis. 



The differences between the two modes of development 

 prove therefore to be as follow: 



(1) The development is more complicated („longer-') in 

 the larval tluin in the Inid development. 



(2) Tlie rudiment starts from another place, from two 

 different parts in the l)ud-rudiment and in the embryo- 

 nal rudiment. 



Great as are tiie diflerences between buds and larvæ, 

 there are points of resemblance to be ftnind hero too. Of 

 these may be mentioned : 



(1) The rudiment of tlie central organs of the nervous 

 system is tubular both in the Ijud and in the larva (the 

 dorsal tube, the medullary tube). (2) This rudiment is, it 

 is true. differentiated in a far sim[)ler manner in the bud 

 than in the larva, the larval brain generally not coming 

 to formation; but those parts of the central organs of the 

 nervous system that are permanent in the full-grown ani- 

 mal, — the permanent ganglion and hypophysis, — are 

 differentiated from the tubular rudiment in the same man- 

 ner in l)oth bud and larval development. (3) The con- 

 nection of the tubular rudiment with the branchial gut 

 (dorsal tubercle) arises both in bud and larva secondarily 

 by coalescence. 



We now tind exactly the sanre laws for the formation 

 of the perihranrliirtl cavity. 



We saw that in the larva, the iieriliranchial cavity 

 originated in 2 dorsal invaginations of the ectoderm. This 

 appeared with certainty from the literature, and from my 

 own personal investigations with regard to the greater 

 numfier of groups; while Pizons different results we found, 

 for the present at any rate. to Ije iiiadniissible on theore- 

 tical grounds. In the buds, however, it appeared that the 

 peribi'anchial cavity was always formed by evagination of 



