♦. 



28G COKTRIDUTTOXS FROM TUK NATIOXAl. TTEHIIAHIUM. 



the soparaio valves (oiitulu iiileriuil Mcf'oinlury casos, nunul or ovnl in rontour, w itli 

 both valves alike or diffi^nMil, in st)]n(^ iiislime(\s hyaliiK^ mid in others variously 

 .seulptured, without i)roeosses or variously oni;uiiented with spines and otherwise. 

 Tln^se internal eases have as a class t]ii(la'r siliceous walls than the external valvi's 

 and girdles whitdi incdose them, and eonscniuently are fretiuently met with in si^adredg- 

 ings, and cspeeiall}' in fossil deposits where iill traces of the (^xlcTnal walls that gav(^ 

 form to the complete^ fdaments have IxM-n lost. As th(>se various parts have bet^n suc- 

 cessively found they have received separat<' g(^n(^ric and spcndfic names, and we there 

 fore have todeal with a tangleof nonienclatunMinusual (^ven among the Diatumaceae. 

 The generic name Jirst to be e(mstiluted and (sul)sequently) apjdied to theae forms is 

 Actinipcus. ]^ecause of its prioj'ity Grunow ^' gives it to the form known (dsewhere iis 

 Bactmastrvm varians T.aud. This nanus however, was niot originally applied to a 

 diatom, Init to a slxdlate silicious si)icule, which Ehrenberg called THvtyocha (Actinls- 

 ens) strius,^ and was not used for the diatoms unlil 1855^ (A. quhmrh(s Ehrenl)/' lacing 

 indi'terminate, but probably also a spicuhO- It is therefore m.t aMiilable fur the 

 Diatomaceao, but must give place to rhatloccrns, 1845. In this c(mnectiun I wish to 

 call attention to a cpiotalion from Van lleurck's Synopsis given below UJub-r Chaetoeeros 

 finratus. In the al)ove list of synonyms *'in ].art'' we find T")I<ladia and Goniothe- 

 cium, bestowed lik(^ Gliaetoeeros and jireceding it in that A^olnnu^ I am, however, 

 opposed to accepting any of these names for ihv. following reasons: (1) They r(^present 

 what are not necessarily structural parts of tlu^ plants, Castracane,^ ludding them to bo 

 si)orangial encasemtuits; (2) Ave ar<^ not even able to affirm that tlu'se spicules are cun- 

 fin(^<l to what we know as the Chactocc^rae; (3) both the above gt^ui-ric names w(Te 

 base<l on fossil fragm(>nts from which we can build no conception of ihe forms to which 

 they belong, and in using them we should, tlui-efore, be nmstituting a genus without 

 any fixed idea of ils structure. Dl( ladia was fiunul in African guario and Goniothe- 

 cium ill diatomaceons (^arth froU) Richmond.ViririTiia./ 



I can not follow rastracane in maintaining the separateness of Ohaetoceros and 

 Bacteriasli-uiii, as his grounds of distinction ,^ aj^pc^ar to bt^ trivial. They are (1) that 

 the valves of J}act(Tias1rum are round and of (diaetoceros generally oval; (2) that the 

 former have mon^ awns or setae than the latter; (3) tliat these are in Bacteriastrnm 

 arranged around tlu^ valve, but in Chaetoceros on opposite sides: (4) that these awns 

 in rhaetocerus interlace. Anycme who will examine a gatluM'ing of Baderiasfnnn 

 rariiDis baud, will see that these characters are in different parts of ih<' same filament 

 too inconstant lo form a geiuw' nmcept. Tims, in II. L. Smilirs type no. 57, both 

 round and oval valvt^s are se(m in lh<' sanu^ fdanu^nt and b(>aring precisely the same 

 kind of setae. Again, the terminal valv<'s often liave two s(4ae, whilo those beh)w 

 Ihem may have more; and in the case of Lauder's figures oi this species /^ Ave fuid 

 terminal valvc^s with eight setae and tlie intermetliate ones with many more. As to 

 the interlacing of \hr setae, it is by no nu'ans c(mstant, many specimens examintnl 

 showing no su(h t(MubMicy. In fact, that condition is harilly to be discovered m 

 Oastracane's own figure.^* 



^^Van Iliun-. Synop. pZ. 82his.J\ 10. ISSI. 



&rhys. Al)h. Akad. AViss. Berl. 1839: Mi)-J5(). 18-11 . Cf. Mikrog. j>L JS.f. r^n-GO. ' 

 ^Ber. Akad. Wiss. Ih^rl. 1854: 237. 1855. 



(^lier. Akad. AViss. lierb 1844: 7(;. 18-15. cf. also GrilT. & Jbnf. Mid'. Diet. ed. W. pL 

 yiS.f. 1-6. 1875, 



eCastr. R(^p. \o\\ diall. 15ot. 2: M. 1880. 



/(T. artich^ by Urighlwoll in Quart. Journ. Mnw Sci. 4: lOn. bS5(i. 

 oCix^Xr. op. cit. 82, 



A Trans. Micr. Soc. bond. n. s, 12: pi. J. 18G4. 

 *'Castr, op. cit. pL UK J. S. 



