[Vol. 9 

 120 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



Algeria: wet places near Bona, about 1860, LeTourneux (Mo. 



Bot. Hard. Herb.); ponds of Chai'ba near Coleah, 18 April, 

 1859, Herb. Font. norm. (Clauson) 99 (Gray Herb.); Dai'a, 

 29 April, 1859, Herb. Font. norm. (Clauson) 99 bis (Gray 

 Herb.); Lac de Ouled Dieb., between Bone and LaCalle, 

 LeTourneux (Gray Herb.); swamps, Boudom, 12 April, 

 1861, Frag. Fl. Alg. Exsic. 499 (Bourlier) (U. S. Nat. Herb.) ; 

 mares de Bou Zegart between Oued Khamis and Djebel 

 Azieb Dahra, Prov. Oran, 20 May, 1875, Warion PL Atl. 

 Scl. (Cosson) 189 (Gray Herb, and U. S. Nat. Herb.); 

 Maisou Cairee, April, 1878, Debeaux (N. Y. Bot. Gard. 

 Herb.) ; without definite locality, 1906, Gandoger (Mo. Bot. 

 Gard. Herb.). 



15a. Forma longissima Pfeiffer, comb. nov. 



/. longissima Bory & Dur. Compt. Rend. Acad 



1844 



gissima A. Br. in Bory & Dur. Expl. Sci. Alg 



fig. 2. 1846-49. 



Differs from the species in the greater leaf length, as much as 

 50-65 cm. in the larger plants. There is a tendency toward a 

 smaller number of leaves (8-12). 

 Distribution : Algeria. 

 Specimens examined: 

 Algeria: LaCalle, May, 1814, Durieu (Mo. Bot. Gard. Herb.); 



LaCalle, May, 1844, Bory (Gray Herb.) ; inundated prairies, 

 24 June, 1814, Motelay (Mo. Bot. Gard. Herb.); LaCalle, 

 186 1, Durieu (Gray Herb.); LaCalle, June, 1884, Debeaux 

 (U. S. Nat. Herb.). 



From the presence of persistent papery brown bases of leaves 

 in some specimens it seems likely that this form may assume 



Engelm 



habit, as in /. Nuttallii A. Br. and / 



The species shows great variability, and its close allies, differ 

 ing from it in sometimes inconspicuous features, are correspond 

 ingly difficult to delimit. Braun and Engelmann were eventualb 

 inclined to consider /. Boryana, I. tcnuissima, I. Perralderiana, . 

 baetica, I. tegulensis, I. dubia and /. longissima as forms or sub 

 species of /. velata. There is doubtless a complex here, whicl 

 properly should be solved by those able to work with an abun 



