40 BRITISH DIATOMACEiE. 



bergii, Kiitz., kindly sent me by M. De Brebisson, undoubtedly belong to tbe 

 next species, I have felt myself obliged to sever this beautiful form from its 

 association with a naturalist who might otherwise worthily claim the honour 

 of giving his name to one of the finest of the Diatomaceae. 



Plate XL. 310. 



3. Diatoma elongatum, Ag. Valve linear, extremities slightly in- 

 flated, rounded, costate ; costee pervious, 18 in '001". Length of 

 frustule -0016" to -0042". Breadth of valve -00015" to -0002". v.v. 



ft. Frustules inflated towards the extremities. Length •0011" to '0025". v.v. 



y. Frustules minute ; valve elliptical. Length -0004" to -0009". v.v. 



h. Frustules distorted, irregular. Length •0004" to "00 11". v.v. 



Ag. Syst. p. 4. Kiitz. Bacill. xvii. 18. Raben. Siissw. Diat. pi. ii. Dia- 

 toma tenue, Rails, Ann. vol. 11. pi. ix. 1. ad specim. authen. in herb. 

 Grev. Var. ft. Diatoma Ehrenbergii, Kiitz. Bacill. xvii. 17. ad specim. 

 quae dedit am. De Brebisson. Bacillaria eJongata, Ehr. Inf. xv. 5. 

 Prit. Anim. iii. 169. Diatoma elongatum, Ralfs, Ann. vol. 11. pi. ix. 2. 

 Hass. Alg. xciv. 3. Var. y. Diatoma tenue, Kiitz. Bacill. xvii. 9, 10. 

 Bacillaria flocculosa, Ehr. Inf. xv. 9. Bacillaria cuneata, Ehr. Inf. 

 xv. 6. 



Fresh or brackish water. Var. a. and/3. Poole Bay, March 1850 ; Lewes 

 March 1852; Denton, Sussex, Nov. 1853; Cork Harbour, x\pril 1855, 

 W. Sm. Belfast, Dr. Dickie. Hornby, Lancashire, May 1853, Mr. G. Smith. 

 Hull, Aug. 1853, Dr. Amott. Tunbridge Wells and Rackham Common, 

 Sussex, Mr. Jenner. Var. y. Tarring Neville, Sussex, Nov. 1853, W. Sm. 

 Ayrshire, March 1854, Dr. Landsborough. Var, $. Hornby, Lancashire, 

 May 1854, Mr. Johnson. SherifFmuir, Dr. Amott, October 1855. (Var. 3. 

 Well near Marseilles, May 1854, W. Sm.) 



This most changeable species has received from observers almost as many 

 names as it presents diversities in size, and yet I feel persuaded, from a wide 

 comparison of specimens, that all the forms I have indicated must be regarded 

 as varieties of the same, dependent upon the stages of its growth, or the 

 circumstances which attend its development. 



Var. ft, which has been the most confidently separated from the type, is 

 perhaps the least constant of any in the character which has been relied upon 

 as a distinction ; frustules with the extremities inflated on a front view, or 

 rather those in which the connecting membrane has slightly collapsed towards 

 the centre of the frustule, being almost invariably found intermixed with 

 others of a perfectly oblong outline, and the two not unfrequently occur 

 together in the same filament. Even the singular variety figured in Plate 

 XLI. 31 1 ■/./', and which Ehrenberg has placed apart as Bacillaria cuneata, 

 will be seen to be accidental, and is found in association with frustules 

 undoubtedly belonging to the normal form. 



