E. M. NELSON ON THE FORMATION OF DIATOM STRUCTURE. 309 



in the areolations near the centre of the diatom. As we advance 

 to the outer zones we find them faint at first and complete at 

 the periphery. We have then a discoid valve, the structure 

 of which is elementary in the centre and complex at the peri- 

 phery. 



Before closing this paper I wish to announce the discovery of 

 the following secondary structures : — 



1. A bar dividing the areolations in Navicula lyra. The 

 bar is parallel to the rapha?. This bar is one of the most 

 delicate of diatomic structures with which I am acquainted, 

 and requires hypercritical microscopy for its resolution. There 

 is also a faint kind of mark between the rows of holes. In the 

 Figure this is a little too dark and a little too broad. Fig. 4. 



2. Minute perforation in Rylandsia biradiata. Fig. 5. 



3. Fig. 6 shows an areolation of Eutogonia Davy ana. 



4. In Coscinodiscus concavns there is an exceedingly delicate 

 perforated membrane. It differs from the Asteromphalus type 

 in not having a circle of larger perforations round the primary 

 areolation. The secondary structure resembles that on Tricera- 

 tium fimbriatum only it is far finer, and, like that diatom, it 

 is best seen from the inside of the valve. That on my own 

 specimen is the finest perforated membrane with which I am 

 acquainted. Fig. 7. 



Thus far these observations are original. 



Fig. 8 shows the secondary structure on Rhabdonema arena- 

 turn. This has been figured before, but the drawing does not 

 agree with my view of the object. 



