M \N N Dl \ rOMS I 'I Til I. A.LB \ I EN >SS VOYAi 2. I 1 



( \,_r. | agree with Castracane thai this union is to be rejected. For, as he points 

 out,° the present genus is never met with grow ingaa Melosira invariably does in long, 

 Bolid, thread-like filaments. \ml although a few species of Melosira present a rough- 

 ened center of the valve slightly resembling the ' umbilicus 3 ' of Podosira, it 

 characteristic of the genus, and where it does occur the blotch is insignificant, 

 shows a But u ml line of separation from the rest of the valve in -lion . is an altogether 

 different thing. So that although individual valves of Melosira and Podosira maybe 

 found somewhat to resemble each other, the typical forms and the modes of growth 



widely different that their union would create much confusion and add uni 

 sarily i" the already unwieldy bulk of Melosira. 1 think the only author following 

 this plan of Smith's is Lagerstedt . 



Kutzing wished to merge Podosira in Cyclotella; but its generally larger size, more 

 delicate structure, convex valves, and, above all, its prominent central umbilicus, 

 make such a classification impossible. 



Ralfs & distinguishes between Podosira and Hyalodiscus by stating that in the latter 

 the vah es are flat. This distinction is not true, and would not be important 

 None of the Bpecies or Variet ies of Hyalodiscus are flat, though in some cases th 

 vexity of the valve is slight. Both have the characterist ic central umbilicus, bounded 

 by an evident Buture. Both show the same Borl of markings, that of the umbilicus 

 being blotched or rugose, that of the rest of the valve being delicate beading, generally 

 arranged in curved lines, producing the appearance of "watch-case milling." Both 

 grow in the same way, one or a few frustules joined by short . Btout, gelatinous stipes. 

 Some species previously classified as Podosira or still so classified by anyone who runs 

 the genus inn* Melosira are wholly destitute of the before-mentioned umbilicus; but 

 that ii has been generally looked upon as an essential mark of Podosira is nol only 

 expressly rtated by Ralfs, bul evident from hundreds of figures of both Podosira and 

 Byalodiscus. Cleve & Grunow get into the difficulty common with many auth - 

 trying to distinguish between the two genera on the basis of the umbilicus. After 

 having stated that " Hyalodiscus hat ein mehr oder weniger scharf abgesondert< 

 brum," ii i- admitted that this is nol universally so, as in varieties of //. subtilis Bail., 

 "so d ii in dieser Hinsicht gar nichl von manchen Formen der /' 



nun ima unterscheiden '." In another place, after stating that Podosira may be looked 

 upon as withoul an umbilicus, the writers add that in the case of /'. nun ima and /'. 

 ambigua "ein ganz entschiedener Umbilicus vorkommt." Tin- whole discussion is a 

 good illustration of the impossibility of holding these two genera separate. Ii i- cer- 

 tainly well known to all who have examined gatherings rich in Podosira thai in the 

 same species and from the same locality the umbilicus varies greatly in si/.- and dis- 

 tinctness, and individuals are not hard to find where scarcely a trace of it remain-'. 

 Thus in the five species of Podosira given in the 11. I.. Smith type, four are with very 

 strong umbilici; one, /'. montagnei Kiitz., a possible Melosira, without. In / 

 moides (Mont.) Kutz. the umbilicus is generally Btrong, in some cases quite small, and 

 in at leasl a dozen valves on this Bingle Btrewn slide almost impossible to see. Van 

 Beurck <* says: 'Th'- Hyalodiscus an- not essentially di fferent from Podosira, except 

 in the umbilicus, which is more or less distinct according to species." 



If. therefore, we remove from this genus the forms thai evidently belong to other 

 genera, chiefly Melosira and Coscinodiscus, we may define it a- follows: 



Frustules growing singly or in a Beries of a few members, attached to a supporl and 

 io the nexl in the Beries by Bhort, Btout, gelatinous stipes, centrally placed these 

 usually causing a pronounced "umbilicus 2 ' or rug situated in the center of 



>. Rep. Voy. 'hall. Bot 2: 139. - 

 /' Pritch. Hist. Enfus. ed. I. Bl I B15. 1881. 



■ Sv. V.i. Aka.l. Ilandl. 17 : II-'. 116. 1880. 

 d Van II. mr. Treat. Dial. W8. 



