18 



nies of which are incrusting or freely ramose there appears in the course of time 

 new layers of zooecia over the old ones, and the old colonies therefore get many- 

 layered. 



According to d'Orbigny the formation of new layers takes place in three diffe- 

 rent manners. In tlie ramose colonies f. inst. in MiiUelea magnifica they are said 

 to start from the proximal part of the colony whence they gradually' and regu- 

 larly extend distally. In the disciform colonies each layer may either f. inst. in 

 SemimuUelea cupula and Sem. gradata be formed by a single subcolony starting 

 from the centre and extending towards the margin or f. inst. in Reptomultelea 

 tuberosa and Clausimultelea tuberosa or the surface of the colonie may at the same 

 time present a greater or smaller number of small disciform sub-colonies which at 

 last must come in contact or fuse together. I have examined a large number of 

 fragments of Mel. magnifica, and I have come to the result that the superficial 

 layers are not formed in such a regular manner as d'Orbigny means, the frag- 

 ments examined presenting in different parts of the surface a number of indepen- 

 dent layers or patches of zooecia. PI. VII, fig. 16, shows a fragment of a colony 

 the one surface of which presents three different layers or sub-colonies. One sur- 

 rounds the proximal part of the rudiment while another arising from the space 

 between the two branches extends both upwards and downwards, and a third, a 

 small round patch is seen to the link side between the two larger ones. The op- 

 posite surface of the fragment presented still two others. While I have not been 

 able to find the ancestrulae of the new layers in Mel. magnifica I have seen a 

 number of them in small fragments of Mel. tubercalata d'Orb. (pi. VI, fig. 3) and 

 Mel. Filiozati n. sp. (pi. VI, fig. 7). As can also be seen in the figure of a young 

 sub-colonie of SemimuUelea gradata given by d'Orbigny such an ancestrula is 

 only represented by the aperture, the rest of the zooecium being covered by the 

 new zooecia which have arisen from it. But while this aperture in d'Orbigny's 

 figure is seen in the centre of a small distinct sub-colony the margin of which is 

 formed by undeveloped zooecia, the named fragments each presents a uniform con- 

 tinuous surface formed by zooecia and heterozooecia among which are seen a 

 number of ancestrular apertures, some of which may often be placed so near to 

 each other that two such apertures are only divided by the breadth of a zooecium. 

 Each aperture which is obliquely ascending is placed in the centre of a small 

 deepening, and the zooecia and heterozooecia surrounding two or more such aper- 

 tures placed near to each other, may be more or less irregularly arranged, but I 

 have never seen such an aperture making the centre of a distinct sub-colony, and 

 the zooecia arising from the different ancestrulae seem to have accommodated them- 

 selves pretty well to each other during their grow'th. A fragment of this species 

 5 "^'"^ long and 3 ™™ broad presents 8 such ancestrular apertures, and another of a 

 similar size 6. A single time I have seen a short cylindrical zooecium placed verti- 

 cally between four zooecia and a kenozooecium, and it must no doubt be regarded 

 as an ancestrula destined to take part in the formation of a new superficial layer. 



