MANN DIATOMS <»r THE VLBATROSS v"OYA< 301 



Biddulphia dubia Bright I leve, in Nordensk. Vega Exped. 3: 508. L883. 1 

 Proc. Acad Phila. 1900: 707. L901. 

 Odontidium punctatum Roper ?, Quart. Journ. Micr. Sci. 7: L80. pi. 9. f. 9. L859. 



Moeb. Diat.-taf. pi IS.}. 9. L8 10 

 / i atium bullosvm Witt, Journ. Mus. Godef. 1: 67. pi. 8.f. . 187 



m dubium Bright. Trans. Micr. So< , Lond. q. 9. 7: 180. pi. 9.f. 12. 

 Schmidt, Atlas pi 



r Biddulphia bicorru Cleve, Bih. Sv.Vet. Akad. Handl. 5 8 : 17. pi. 5. 

 /'. . L878 

 Ampkitetras bico nis De Toni, Syll. Alg. 2: 902. L894. 

 Though this species has sometimes a rather close resemblance to minute specimi 



data Eloper, and this led me for a time to make it a variety of the latter, a careful 

 examination of many specimens of both forms, abundantly supplied in some of the 

 dr edgings, shows them to be essentially distinct. B. dubia is always small, its reticula- 

 tions irregular and of unequal size, their dividing walls thick and crinkled, the border 

 massive, and the valve outline rhombic. In/;, etic lat a, on the other hand, a 

 diatom, the reticulations are quite regular, generally hexagonal, with thin dividing 

 walls showing "knots 6 ' at their points of juncture, the border is nol so massive, and the 

 valve outline is either elliptical or with convex sides approaching that figure, 

 species have secondary dotted nun-kin::- within the reticulations, but thosi IB 

 lata are smaller and more distincl . The general similarity of these two Bpecies has, I 

 think, led to the naming of Schmidt's figures 6 '/; eticulata Roper, var.?" whereas 

 they are certainly closer to B. dubia. The union of 7Vw*< atium bict CI e, and 

 itium bullosum Witt, which i§ recognized by I)'- Tonic and by Boyer,<* is rather 

 difficult to admit, in view of Witt's carefully drawn figure and his plain description. 

 Yet the close similarity of Witt'sform t m dubium Bright, is undeniable, as 



is also the specific identity of T. dubium and '/'. In fact, we have here an 



illustration of how misty and indefinable our specific boundaries really are. and how 

 subsequently discovered specimens may bridge over the wide gap thai 

 apparently quite unlike species. This same condition has been previously referred to 

 under Auliscus punctatus and .1. p Bail. 



Found at station 2885, off Oregon. 



Biddulphia edwardsii Febiger; IT. L. Smith. Sp. Diat. Typ. no. 623. L874. Boyer, 

 Proc. Acad. Phila. 1900: 701. L901. 



Odontella edwardsii Grun.; Van Ilenr. Synop. pi. 100. f. 9- 10. 1881. Grun. Denkschr. 

 Akad. Wien48 2 : 57. pi. :./. 20. L884. De Toni, Syll. Alg. 2: - - 



Biddulphia obtusa Ralfs, err. det. in Van II. air. Synop. pi. 100. f. :: ;:. L881. 



Biddulphia polycanihos Brun. Mem. Soc. Phys. el Bist. Nat. Geneva 31 1 : L2. />!. l .'. 

 i-b. L891. 



Odoitt<Ua ? polycanthos De Toni. Syll. Alg. 2: 865. L894. 



I find two well-marked varieties of the above polymorphic diatom. One < orr< sponds 

 closery t" what Brun 1"<-. « it. ha- called*/?, polycanthos, and that, t<">. in tin- forms 

 both witli ami without the large spines. Brun therefore erroneously makes these 

 spurns a mark of specific distinction for his form. Nor can the difference of /;. "/ 

 having diagonal and /;. polycanihos vertical lines of beading on the girdle, be admitted. 

 For in most gatherings i i" B. edwardsii, where there is considerable diversity in size, 

 both patterns of marking can be Been in abundance. Such, for example, is the ease in 

 the II. L. Smith type-slide no. 623 in possession of the U. S. National Museum. This 



I • • Toni, Syll. Alg. 2: 641. L892 

 6 Schmidt, Atlas pi. 78. f. 

 < I).- Toni, Syll. Alg. 2: 973. L894. 

 d Proc. Acad. Phila. 1900: 707. 1901. 



