KONGL. SV. VET. AKADEMIENS HANDLINGAR. BAND 25. N:0 5. 137 



Long. 26, lat. s. acul. 20, lat. isth. 6.5, long. acul. 3-4 t u. T. XII, f. 9, after 

 Wallich Mscr. No. 192, f. 1. 



13. A. iiicavatus, n. sp. A. minimus; semicellula? elliptico-fusiformes, dorso rotundatas, 

 apiee rotande ineavata?, e basi rotundo-inflata?, angulis lateralibus sub-acutis, mucronibus 

 brevibus 4 instructis; membrana glabra; incisura mediana lata aperta interne rotundata. 

 A vertice visus ovalis, apicibus acutis. 



Long. 10, lat. 12, lat. isth. 4, crass. 5, long. mucr. 0.5 fi. T. XI, f. 43. 



Xearcst to A. tenuissimus Archer (Micr. J. Sci. p. 175, t. VI, f. 50 — 55, 1864), 

 but is a little larger and bas onty 4 mucros. none being dorsal. I have always re- 

 garded Archer* little Irish plant (sinee found in Sweden, Norway, England &c.) as a 

 Xanthidium (tenuissimuin) on aecount of its having two pairs of dorsal mucros. In 

 the same way I wotild class Boldts form (Sibr. Chlor. t. V, f. 16) as Xanth. hexa- 

 gonura (Boldt) Nob., whilst I would retain Boldts name for the other form (f. 17), 

 which in spite of its central inflation is in all other points correctly described as 

 Arths. hexagonus by its author. 



14. A. morsus, n. sp. A. minor, paullo latior quam longus; semicellulis indistincte hexa- 

 gonis, dorso protentis sub-acute incisis, angulis basi rotundatis superioribns plus minus 

 acuto-rotundatis; incisura mediana magna rotundata non profunda; membrana glabra; 



. processubus 4 crassis glabris rectis parallelis apiee rotundatis lateraliter positis. Cor- 

 ])us amvlaceum in qtiaque semicellula singulum. A vertice visus medio late ellipticus 

 fere circularis. 



Long. 20, lat. 23, lat. isth. 6.5, long. proc. 6 u. T. XVI, f. 33 (inter Staurastra). 



This little form seems verv anomalous on aecount of the processes being so 

 thick and not attenuate. The nearest species is A. pachycerus Lagrrh. (Amer. Desm. 

 Flor. p. 244, t. XXVII, f. 23) from which it dififers in the incised apices and the non- 

 divergent thicker arms, the vert. view is also broader in proportion. It is also like a 

 small thick 2-angled form of S. globulatum, Bhéb. 



15. A. gibbernlus, Joshua (on Desm., J. Bot. p. 33, t. 254, f. 6, 188,">; Burm. Desm. 

 p. 644). 



Long. 30—35, lat. s. acul. 26—29, c. acul. 52—58, lat. isth. 9—11, long. acul. 

 12—14 /ii. T. XII, f. 5. 

 Central India. 



I do not think the genus Arthrodesmus can be more convenientlv divided than 

 by following the suggestions of Archer, in Prit. Inf. pp. 736 — 7, 1861; the sub- 

 genera I propose are: 



A. Aplodesnms (aziÅoog, simplex; d€G,uog, vinculum). Angulis lateralibus integris. 

 Typ. sp. Arth. convergens, triangularis, subulatus, gibberulus. 



B. Schizodesmus (öj/sW, scindo; &c). Angulis lateralibus fissis. 

 Typ. sp. Arth. bifidus. 



K. Sv. Vet. Akiid. Haudl. Hand 25. N:o 5. 18 



