332 



DISCOVERY REPORTS 



growing and thickening wall of the float that the first gastrozooid's relationships become obscured in 

 later growth stages (Text-fig. 1 2). The main foramen is that of the peduncle of the tentacle of group 2. 

 (2) and (3) Tripartite groups * first and laterals. The general pattern of arrangement of the tripartite 

 groups is easy to recognize once it becomes familiar (PI. XIV, figs. 1-4). These groups were first des- 

 cribed by Schneider (1898), who had only two specimens to work on, under the name of ' Urgruppen ', 



T, i=J_ V LEEWARD 



\ 



I 



MM 



LEE BRANCH 



GON 



GASTROZOOIDS 



GONODENDRA t!£"-"v3 



LEE BRANCH 



Text-fig. 14. Physalia physalis. Two views of cormidium VI (main zone) of a young left-handed specimen, number 11, to 

 show stages in the growth of the lateral groups and lee-branches. A = oral view; B = aboral view, x 9-5. 1st group with- 

 out a tentacle, GZ 1 and GON 1 ; lateral groups, gastrozooids GZ 2-5, gonodendra GON 2-5, tentacles T 2-5. 



and although his account is difficult to follow without figures, it is a sound analysis of the tripartite groups 

 and their subsidiary lateral branches. Each 'Urgruppe' (Text-fig. 15) consists of one polyp (gastro- 

 zooid), one tentacle and one genital cluster (gonodendron), all borne on a single peduncle (PI. XIV, 

 fig. 1). These simple 'Urgruppen' very soon become complicated, said Schneider, by the develop- 

 ment of second, third and fourth ' Urgruppen ' on similar peduncles, each arranged on one side of a 

 thickened peduncle and separated from one another by lateral peduncles. Here he was probably 

 referring to the relationship between the first tripartite group of a cormidium and the series (3) which 



* The significance of these is explained on page 361. 



