UMBELLULIDAE 259 



Antarctic waters together with U. magniflora Kolliker 1880, and the 'Valdivia' Expedition brought 

 back a third species, U. antarctica Kiikenthal 1902, from the same area. Two species from the Indian 

 Ocean were also erected in 1902 by Kiikenthal, U. rigida and U. pellucida, and several other species 

 have been added since that date. 



Both Jungersen (1907) and Hickson (1907) expressed their doubts as to the validity of the three 

 antarctic species carpenter!, magniflora and antarctica. Jungersen who gave careful details concerning 

 U. carpenteri, wrote (1907, p. 9): 'es kommt mir deshalb als sehr moglich vor, dass diese Art nicht 

 nur noch grosser wachsen, sondern hochst wahrscheinlich die Zahl der Polypen bedeutend vermehren 

 und die ganze Form des Schopfes andern mag, mit anderen Worten sich den Typus der U. encrinus 

 (oder deren antarktischen Vertreters der U. magniflora=U. encrinus var. antarctica Kkth.) mehr 

 naheren kann'. In 1937 Hickson was inclined to go yet a step further, and discard the shape of the 

 axis as a specific criterion, considering all spicule-free Umbellulae as one species. 



It is appropriate to consider here the distinguishing characters which have been used in this group. 

 In his key Kiikenthal (1915) gave them as follows: 



jPolypentrager kurz, Polypen annahernd in konzentrischen Kreisen stehend. 2 



iPolypentrager lang, Polypen bilateral angeordnet. 11 



2. Achse vierkantig. 3 



3. Polypen ohne Spicula. 4 



(Tentakel ebenso lang oder langer als der Polypenkorper. S 



(Tentakel kiirzer als der Polypenkorper. 8 



I Stock elastisch. 6 



I Stock Starr. 7 



, ( Kiel ausserst diinn. U. lindahli 



■ I Kiel relativ dick. U- encrinus 



fSchopfkelch lateral abgeplattet. U. carpenteri 



I Schopfkelch dorsoventral abgeplattet. U. rigida 



(Polypen zahlreich in mehreren konzentrischen Kreisen 9 



(Polypen in geringer Zahl in einem Kreise. U- magniflora 



(Polypen schlank und sehr gross. U- Antarctica 



I Polypen dick, tonnenformig und klein. ^- pelluada 



11. Achse vierkantig. '^ 



12. Polypen ohne Spicula. '3 



(Tentakel lang. U. purpurea 



^^' (Tentakel kurz. U. kollikeri 



A glance at this key immediately raises the question: how to distinguish between the terms 'short' 

 and 'long'? None of the species mentioned in it has a rachis comparable with that of U. spicata, the 

 only one which can be characterized as 'long'. It is also questionable whether a difference exists 

 between arrangement in approximately concentric whorls and bilateral arrangement. The young 

 specimens in most species are more or less bilateral ; the older specimens generally have fairly distinct 

 whorls, concentrically arranged, sometimes indistinctly bilateral. 



The other characters of the key in most cases are questions of degree or gradation, or are due to 

 different stages of contraction. Some (e.g. point 5, whether the stem is rigid or elastic) might seem 

 legitimate ; however, the more specimens we investigate, the more difficult are the cases we run up 

 against. Specimens are always more flexible when first caught than later on, and, of course, dry they 

 are always much more rigid and brittle than if wet ; in formalin or spirit their condition is intermediate 

 between these two extremes. Rigidity is correlated to a certain degree with calcareous encrustation, 

 with size of the specimens, etc. It must also be emphasized that flexibility evidently increases to some 

 extent with the thickness of the sheathing tissue, as can be observed in typical colonies of 



