DESCRIPTION OF SPECIES 73 



have a pointed triangular languet. In the terminal zooids only, that is those immediately round the 

 common cloacal opening, the end of the atrial languet is wide and straight with a row of pointed or 

 leaf-like teeth. These teeth correspond to the teeth present on the test round the common cloacal 

 opening, already mentioned, and doubtless are responsible for their formation. 



The intestine only occasionally shows the subdivisions which Kott (1954) used to distinguish 

 S. sigillinoides from S. quoyi. 



Larva. The larvae from different colonies vary in size from 0-40 to 0-76 mm., measured from the 

 end of the papillae to the base of the tail, the majority being between 0-50 and o-6o mm. Caullery 

 (1909), Salfi (1925) and Kott (1954) have already described the larva. The sensory vesicle contains 

 only one black pigment spot, the spherical otolith, 22-26 /i in diameter. 



Biology. Breeding probably does not occur in the southern winter, yet as shown by the figures in 

 Table 22 it is not restricted to a short period of the year. 



Month May 



No. of colonies examined 3 



No. of colonies with 1 



embryos 



In different colonies the embryos in the brood pouches numbered from two to fourteen. 



The heads and zooids are unisexual, but all zooids in a head are not at the same stage of develop- 

 ment. In a single head buds and young non-functional zooids may be found at the base ; then a zone 

 of functional but sexually unripe zooids occurs; and finally the distal part of the head is devoid of 

 zooids but contains many isolated brood pouches with embryos and larvae. The brood pouches dis- 

 appear finally, having presumably been set free by the disintegration of the common test. Whether 

 the larvae escape before or after this is unknown. 



It has long been known that heads occur free in the water, and indeed the first record of the species 

 (Lesson, 1830) was of an isolated pelagic head. Free heads were frequent amongst the 'Discovery' 

 material, having been taken in various years in January, February, October and December, i.e. in the 

 southern spring and summer. Brewin (1953) suggests that the breaking off of heads is caused by rough 

 weather, but van Name (1945) considers it as probably a normal event in the life of the species. Head- 

 less stalks, at any rate, seem to be a part of the annual cycle (PI. II, fig. 5) ; they occur in the ' Discovery ' 

 collections sometimes in December, but most often in May, June and July, i.e. in the southern winter. 

 At the depths from which these headless stalks were collected (137-304 m.) storm action could not 

 account for the loss of the heads, and I conclude that they were lost in the normal course of events. 

 But headless stalks do not necessarily imply free pelagic heads, as heads disintegrate after breeding. 

 Whether intact heads also break off, except by storm or other accidental damage, is still doubtful. 



The headless stalks do not die, and PI. II, fig. 7, illustrates the growth of new heads. The old stalk 

 consists of a firm outer layer with a sharply defined upper margin, and a softer central core of test 

 material containing many small rounded orange bodies and perhaps a few remaining vascular processes 

 from the zooids which formerly occupied the head. Caullery (1909) and Salfi (1925, 1926) have shown 

 that these rounded bodies are buds formed by multiple constriction of the vascular processes of the 

 zooids. The reduced state of the colonies, containing the buds but no zooids, recalls the overwintering 

 condition of colonies of the Family Polyclinidae, but I do not know if the colony of Sycozoa also has 

 a period of quiescence. When the buds develop they produce new zooids within the central test near 

 the upper end of the old stalk, and at the same time new central test material begins to bulge out 

 from the end of the stalk. As the new test continues to grow it forms a new head containing the 



