21 



Botryllidæ indhffattei- de 2 slægter: 

 Symplegma, Heidman, 188G. 

 Botryllus. GocitiKT, 1774. 



Til (k-n første liorcr lilot cii eneste art tra Berjnndas 

 øerne. beskreven af Herdnian. Den anden sUegts farverig- 

 doni liar opfoidret til opstillen af subgenera, hvilke har 

 frt'inlni(lt store vanskelioheder j)aa grand af arternes con- 

 formitet. Man har derfor forelobig endog maattet noie sig 

 med at benytte koloniens tykkelse som karaktermerke: hvad 

 der er ovei' og under 5'"'", et maal. der er ganske vilkaarlig 

 valgt. Heller ikke er tykkelsen videre constant. Tnge 

 conni af poJ/jc//chi.<t eller scu-cobofri/lhi/des er saaledes under 

 det opstillede lavmaal 5""". Slægtsbesteinuielseii kræver 

 derfor undersøgelse af et større antal kolonier og helst til 

 forskjellig aarstid. naar man da ikke er saa heldig straks 

 at finde en tvk cornius. 



Alinindelig opstilles 4 folgende subgenera (Di-asche, 

 Herdman. Lahille): 



Systemerne 

 hovedsagelig 

 kredsformige 

 el. elliptiske. 



Systemerne 



hovedsagelig 



åren ed e 



Boivj/lhiK (Goert.). Tykkelsen af coiMnns gaar 

 aldrig over 5""". 



Poli/CjiciHx (Lam.). Tykkelsen af cormus over- 

 stiger altid o'""". 



Bofri/llnJiJcs (M.-Edw.). Tykkelsen af cormus 

 gaar aldrig over 5""". 



SaiTohotrijlloides (Dr.). Tykkelsen af cormus 

 overstiser altid o""". 



Lahille vil kun bibeholde disse slægter som rent jiro- 

 visoriske, idet han venter, at der ved mei-e omfattende un- 

 dersogelser vil fremdukke sikrere anatomiske eller blasto- 

 genetiske slægtskarakterer. Han uiver et tingei-])eg for 

 classificationen ved sm omhyggelige beskrivelse af tentakler 

 og branchialsæk untler sin behandling af shegten jjo///r!/r/»,s. 

 Herdman stiller tentaklernes antal og foi'm i forste nekke 

 ved artsbestemmelsen. Hos de faatallige arter, jeg har 

 haft anledning til at undersoge. har jeg fundet disse 2 ka- 

 rakterei' at være de paalideligste. For tentaklernes ved- 

 komniiMidc maa man dog forsikre sig om at have fuldt ud- 

 viklede individei- for sig. da man ellers resikerer at tinde 

 et for ringe antal. Antal stigmata i branchials:ekken er 

 heller ikke fuklt nd constant. idet hos hver art l)aade an- 

 tallet af de verticale i'ækker varierer (f ex. fra II) — 12), 

 ligesoin formlen for hver af disse igjen er noget vexlende, 

 begge dele endog inden samme cormus. Man tinder saaledes 

 hos en art alniindeligst f ex. folgende foiMnel: d. (dorsal- 

 side) 8 — 5 — 5 — 8 r. (ventralside); de 3 streger betegner de 

 3 paa branchialsækkens indvendige side longitudinelt lobende 

 bjælker (cotes longitiulinales, Lahille); ved siden lieraf fore- 

 kommer ogsaa formlerne: d. 8 — 4 — 4 — 8 r. og d. 8—5 — 6 

 — 9 r. Ved branchialsækkens beskrivelse kan derfor blot 

 anføres grænserne for eller det almimlcligst forekom- 

 mende antal af stigmata. Angaaende grupperne hotri/l- 

 liis (imlycydus) og hotrylloides (mrc<)J)otrylloides) relationer 

 til hinanden med hensvn til branchialsækkens dannelse, saa 



The faniily Botryllidæ includes the two genera: 

 Symplegma. Herdman. 188fi. and 

 Botryllus. Goertner, 1774. 



()nly one species. fi-oni the Rermudas, aud descrilied 

 by Herdnian. belongs to the first of these. The wealth of 

 forms in the second genus has called for the establishment 

 of sub-genera, and this has presented great difticulties on 

 acconnt of the couformity of the species. It has therclbri' 

 been Ibund necessary to employ, for the jireseiit, the thiek- 

 ness of the colonies as a characteristic mark, distinguishing 

 between those that are over. and those that are under 5""". 

 a ineasurement which is arbitrarily chosen. But neither is 

 the thiekness very constant. Young eormi of Poh/cychis 

 or SarcohotrylJoides. for instance, are under the given 

 minimum nicasui'ement. 5'""'. The determination of the 

 genera requires, theiefore, the exaniination of a large num- 

 ber of colonies, and if jiossible at different seasons of the 

 year, if a thick cormus docs not chance to be found im- 

 mediately. 



The four following sub-genera are grnerally recoguized 

 (Drasche. Herdman. Lahille): 



Systems 



chiefly cir- 



cular or 



elliptical 



Systems 



chiefiy 



raniified 



liotrylhis (Goert.). Thiekness of c(M-nius never 



exceeding 5'"'". 

 Polycyidus (Lam.). Thiekness of cormus al- 



ways exceeding 5""". 

 Bofrylloides (M.-Edw.). Thiekness of cormus 



nevei' exceeding 5""". 

 Sarcobofrylloides (Dr.). Thiekness of coi-mus 



always exceeding 5""". 



Lahille only retains these g'Miera as quite provisional, 

 expecting that more comprehensive investigation will bring 

 to liglit more ceitain anatomical or blastoiienetic generic 

 characters. He gives a hint foi- classitication in his careful 

 descri[)ti(ni of tentacles and branehi.-il sac in treating ni' 

 the genus Polycydus. Herdman c(Hisidei-s the numl)er and 

 form of the tentacles of first im])ortance in the determination 

 of the species. In the few species that I have bad the 

 op])ortunity of examining, I have found these two charac- 

 ters to lie the most reliable. As far as the tentacle.s are 

 coucerned. one iu\ist be assured of having a fully developed 

 animal to examine, as otherwise thei'e would be a risk of 

 finding too few. Neither is the nunil)er of stigmata in the 

 l)rauchial sae peilectly constant. as in eacli s])eeies the 

 nuniber of vertical rows varies (e. g. Iroin Kl to 12) and 

 the formula for each of tiiese again is rather variable even 

 in the sanu' cormus. Fn one .species, foi- instance, tiie 

 following formula is found to be the most general: d (dorsal 

 side) 8 — 5—5 — 8 r. (ventral side). The 3 dashes signify 

 the 3 longitudinal bars (cotes longitudinales. Lahille) on 

 the branehial sac. Beside this oceur the formulæ (/. 8 — 4 

 — 4—8 v. and d. 8 — 5 — 6 — 9 r. In describing the bran- 

 ehial sac. therefore, only the limits, or the most tVe(|uently 

 occurrin;j; number of stigmata are here given. With regard 

 to the relative formation of the branehial sac in the groups 

 BotryUu.^ ( Polyydus) and Botrylloidcti (Sarcobofrylloides), 

 1 may state that, according to my experience, the number 



