294 PROCEEDINGS OF SECTION D. 



aculeatimi, I.e., 1910, PL ii, f. 17), and then by degrees into smooth 

 bifid processes (var. hifarhim, PI. ii, f. 18), which grow out to a 

 definite length, var. Tohopekaligense (Wolle) mihi (cf. W. & G. S. 

 West, Frw. Alg. Ceylon, PI. 21, f. 27). In all the specimens I have 

 seen, and these forms are fairly general round Sydney, the lateral 

 processes grow out one corner at a time (the general rule indeed 

 with all processes). First a delicate spine shows alone, then a 

 denticulation forms a httle above and pushes out, carrying the 

 intervening membrane with it, thus forming a bifid process. Later 

 on a couple more spines often grow out from the angle between the 

 first two ; this is quite the general thing in Staurastrum. 



I cannot see why there should be any difficulty in accepting 

 the growth of spines on the cell, seeing that it is well understood 

 that much more elaborate ones develop by a regular process on 

 the zygospore. 



Absorption. — Not only do granules and spines develop, but 

 they are occasionally absorbed again. A good example of this is 

 Xan. simplicius, Nord. In var. hotanicum, Playf. (I.e., 1908, 

 PI. xi, f. 10, p. 620), which is a young form, there is a long spine 

 at each angle. As the cell matures, however, the body of the semi- 

 cell grows outwards, absorbing part of the spines, and a second 

 spine forms, first at the upper and later also at the lower angles. 

 In the young form the spines are about 9m long, in the type only 

 half that length. Again, in Siaurastrum the granulation in young 

 forms is very often coarser (the granules larger, fewer and more 

 distinct) than in the more fully developed state. This occurs in 

 St. piinckdatiim and St. hexacerum sometimes ; the granules are 

 probably partly absorbed and fresh intermediate ones formed. 

 The same process has already been mentioned with regard to the 

 striolation of Closierium, partial absorption of costae taking place, 

 with simultaneous formation of intermediate striae. It should be 

 remembered that spines and processes are hollow, even those most 

 delicate spicules that sometimes adorn the apex of the verrucae in 

 St. assurgens and Cos. subspeciosmn, and that they are in vital 

 connection with the contents of the cell, even up to the very tip, 

 by means of a minute thread of homogeneous protoplasm. The 

 same is true of the setaceous rostra of certain Closteria. W. & G. S. 

 West have indeed described the processes of St. Freemanii {Frw. 

 Alg. Ceylon, p. 177) as being solid, but this apparent solidity I have 

 reason to believe to be an optical illusion. It is another instance 

 of the appearance of incrassation already remarked on. The 

 illusion is caused by the hollow process being filled with homo- 

 geneous protoplasm of a highly refractive nature Every granule, 

 tooth, spine, and process, however small or however elaborate, is a 

 hollow prolongation of the cell- wall. 



American workers in this field have the opportunity of a very 

 interesting investigation into the polymorphism of a species and 

 the growth of spines and development of the cell in the life-history 



