22 W. M. BALE ON .SOME OF THE DISCOID DIATOMS. 



in C. gigas, G. diorama, and a few others, it is the central part 

 of the valve which is so modified. Though in C. apiculatus and 

 C. perforat us it is universally recognised that this peculiarity is 

 not of specific importance, the loose disposition of the markings 

 in the central part of such species as C. diorama has been made 

 use of to characterise the species, but in some cases at least 

 unwarrantably. In a species found in Port Phillip the larger 

 valves have the markings as in G. diorama, while the smaller 

 ones are areolate throughout. When the modification in question 

 occurs in the central part of a valve it is usually associated with 

 a thinner condition of the silex, but this does not appear to be 

 the case in such species as G. perforatus and G. apiculatus. 



In rare cases the loosely disposed and rounded markings occur 

 on an annular area, concentric with the margin, and an interest- 

 ing example of this is found in the large, robust form of G. Oculus 

 Iridis found in the Mors deposit. It is a variable form as regards 

 the surface contour, but commonly in large valves the centre and 

 the sub-marginal zone are about equally elevated, and the inter- 

 vening broad annular area is slightly depressed. A varietal 

 form differs in having this depression much deeper, and, on the 

 outer side, very abrupt, while in a third form the annular 

 depression is very deep and narrow, and on the bottom of the 

 depression the cellules are rounded and separate (a condition to 

 which there is sometimes a tendency in the second form). This 

 last variety was described by Grunow in his work on the diatoms 

 of Franz-Josef Land as a new species, under the name of 

 G. annidatus, notwithstanding which it was figured later on 

 PI. 184 of Schmidt's Atlas under the name of Craspedodiscus 

 Molleri. 



I have also seen a form of G. excavatus, very near to Grunow' s 

 var. semilunaris, in which there is a complete annular depression, 

 with round markings, not far from the centre. 



The circular areas of the varieties just mentioned, as well as 

 the inflations of ordinary forms of G. excavatus, are all instances 

 of abrupt bulging in (or out) of the substance of the valve, and 

 in all of them the portion which is subject to this bulging 

 appears thinner than the rest of the valve, while the markings 

 are fainter, as well as being rounded and loosely disposed. 



The occurrence of " bright points " at the origin of the shorter 

 radial series of cellules has been commonly regarded as a valid 



