858 Transactions. — Botany. 



I found this Diatom in quantities during the months of April and May, 

 in a spring at Ngapari on the side of North Moeraki Downs, facing the river 

 Ashley. 



I am indebted to Professor Hutton for the verification of this descrip- 

 tion. 



Plate XXII., fig. 1 : a, front view; b, suture; c, side view. 



On the Identity of Amphicampa xvith Himantidium, and Description of a neiv 



Species. 

 {Read before the Philosophical Institute of Canterbury, 4i7i August, 1881.] 

 About four years ago I found, quite accidentally, in a di-ain in the Oust 

 Valley swamp a small pocket of diatomaceous earth, which on examination 

 much resembled Amphicampa mirahilis, Ehr., described in the Micrographical 

 Dictionary as a doubtful genus of fossil Diatomacea. The other species 

 named is A. eruca, and both are represented as found fossil at Tizar, 

 Mexico. 



Professor Hutton found fossil valves in Waikato and elsewhere, and 

 Mr. George Gray recognized them in gatherings taken from the Eiver 

 Avon. 



In the month of May last I gathered a considerable quantity of a 

 filamentous Diatom from a spring at Ngapari, Fernside, which, after treat- 

 ing in various ways, I resolved into the hitherto so-called genus Amphicampa 

 (both sides sinuated). That genus, however, appears to have been con- 

 sidered " free," while in the form which I gathered it is filamentous, and 

 cannot be separated generically from Himantidium. The present species is, 

 however, certainly not A. eruca ; and as A. mirahilis is figured with f bends, 

 while my species varies from ^ to f, I assume it to be a new species. 

 Himantidium maskellii, sp. nov. 



Frustule : front view rectangular ; length about four times the breadth, 

 straight, forming a lengthened tenacious filament, showing lines of dots at 

 the points of sutm'e corresponding to the striae on the side view. Valve : 

 elongated, shghtly arcuated, and attenuated towards the extremities, which 

 are boldly rounded ; both edges are sinuated ; length about seven times 

 the breadth. The convex edge has invariably one more bend than the con- 

 cave. The valves vary in the number of the bends from f to f . Stri« : 

 parallel and transverse, and there are 23 to 27 to the -001 of an inch. 



Hab. — Ngapari, Fernside, and Eiver Avon, North Canterbury ; Wai- 

 kato and Cabbage Tree Swamp, Auckland (fossil) — Hutton; Cust Valley 

 Swamp, North Canterbury (fossil). 



Plate XXII., fig. 2 : a, front view filamentous ; c, side view. 



