AN ALGAL LIMESTONE FROM ANGOLA. 583 



12m x 16/*. The outline of the cells in some of the sections is polygonal, often with 

 five sides. Some of the tangential sections show large polygonal markings ; these 

 are clue to growth having taken place from different adjoining centres, with the 

 resulting compressional structure (Plate). 



The most characteristic feature of the species is the presence of reproductive 

 organs, which have been observed in several sections, and in sections made from one 

 specimen they were very abundant. 



The conceptacles have a distinctive shape, as will be seen from the plate. Their 

 dimensions bear a close resemblance to those of species which are mentioned by 

 Dr Rothpletz in his paper on " Calcareous Algae." * In L. ramosissimum Reuss. the 

 measurements for the conceptacles are given as 280//, long x 80/x high, in L. suganum 

 Rothp. as 250/a long x 100/x high, and in L. torulosum Glint, as 400/x long x 150/x. 

 high. From the plate it will be seen that owing to the curved shape of the 

 conceptacles measurement is not satisfactory, but an average of well-formed ex- 

 amples, reckoning from edge to edge, gave 350/x long x 125/x high. The conceptacles 

 occur most usually towards the edge of the thallus. Although they are usually 

 isolated and are situated most commonly on the same circumference, it seems that 

 sometimes they may be so close together as to coalesce in forming a band of " con- 

 ceptacle tissue," with the loss of individual shape. In a single thallus, isolated 

 conceptacles can be seen to pass into the conceptacle band. No spores have been 

 observed in the conceptacles, although they contain a certain amount of structure, 

 suggesting canals. When writing on the classification of the Nelobesiese, Foslie f 

 distinguishes between those Lithothamnia in which the sporangia are grouped in 

 conceptacles, and those in which they are not, placing the former in subgenus 

 Eulithothamnion. This subgenus he divided into two sections according to 

 whether the conceptacles remained superficial or not. In the species under dis- 

 cussion the material of the thallus subsequently grows round the conceptacle, so 

 that this species falls into the section Innatse Fost., to which section also belono- 

 L. ramosissimum, L. suganum, L. torulosum. The conceptacles in the first two are 

 smaller, and in the third larger, than in this species. It resembles those three which 

 are Tertiary most nearly, but is distinguished from them by the characteristic shape 

 of the conceptacles. The specific name of Lithothamnion angolense, n. sp., is pro- 

 posed for it. 



Girvanella forms the largest proportion of the nodules, but the two have been 

 observed in association in all the sections examined. It appears that bundles of 

 Girvanella tubes have first been formed, enclosing fragments of echinoid plates or 

 quartz grains, and that subsequently Lithothamnion has begun to grow on the 

 nodules so produced. Its growth has, however, been arrested at an early stage, for 

 in no instances are large masses found, and the original outlines of the Girvanella 

 nodules are only slightly modified. 



* A. Rothpletz, op. cit. t H. Foslie, Kongl. Norske Vidensh Selsk. Skr., 1900, No. 5, p. 12. 



