187 



ATLAS DER DIATOMACEEN KUNDE. 



Since our last notice of this work two more parts have appeared. 

 In part II., a further instalment of the fastuosa group of Smirella; 

 the 5th plate contains fifteen figures, ten of which are devoted to 

 Surirella fastuosa and its varieties. The author promises' to give 

 very shortly a plate containing some splendid new varieties of this 

 group. 



In plate 6 we return to the Naviculce. This plate contains 41 

 figures of Navicula humerosa, and the more or less allied forms. It 

 is probable that several of these forms have been previously pub- 

 lished under other specific names, e.g., Navicula Fisclwri is Stau- 

 7'oneis maadata, Bailey's Microscopical Organisms, p. 40, pi. 2, fig. 

 32. Pritchard's Infusoria, 1864, page 915 ^ N. Johnsoniana, Gre- 

 ville. Transactions of Microscopical Society, vol. xi., p. 17, pi. 1, fig. 8. 

 Plates 7 and 8 contain 116 figures of the elliptiform Naviculce. N. 

 gemmata, var. spectahilis, Grunow, seems to be identical with N. 

 gemmata, var. biseriata of the same author, a figure of which was 

 given in vol. i. of Grevillea (see Novara Diatoms, p. 91). This 

 part contains 172 figures. 



Part III. commences with the Cymhellce, of which plate ix. con- 

 tains 79 figures. This very puzzling genus is admirably represented. 

 Its close alliance with the Naviculce can be easily traced from the 

 conspicuously lunate valve and eccentric median line as exemplified 

 in C. gast}-oides, to the symmetrical sided C. moUeriana, in which 

 also the median line and nodule are perfectly central. 



Plate 10 contains 69 figures of forms closely allied to the pre- 

 ceding genus, and which perhaps ought to be united with it, viz., 

 Cocconema and Encyonema. Fig. 36 represents a very curious 

 abnormal form of C. mexicanwn, the median line becoming 

 double on one side of the central nodule, the space between the 

 double line being punctate striate, like the other part of the valve. 



Figs. 42 to 69 represent species of the genus Encyonema. 



Plates 11 and 12 give 105 figures of the panduriform Naviculce 

 ( = Dijiloneis, Ehr.). We are scarcely prepared to admit that 

 specific differences exist between iV. Panclura, N. separahilis and 

 N. 7nulticostata, and careful comparison of figui-es 1 to 12 and 14 

 to 20, pi. 11, will show a gradual merging of these forms into each 

 other; in fact, the author remarks in the explanation of this plate 

 that N. separabilis stands between N. Pandura andiV. multicostata, 

 the former approaching the last very closely, and that Grunow 

 makes it var. intermedia of his N. multicostata. 



Plate 12 commences with 12 figures of N. interi'iipta, 5 of un- 

 named species, q,nd 56 of vai'ious new and previously described 

 forms. The figure of Navicula giebelii is much inferior to that in 

 the specimen plate. This group will be continued in next number. 

 The description of the plates in Part III. is much less convenient 

 than that in Parts I. and II. ; instead of placing the generic and 



