Prof, de Notaris on Ginnania furcellata. 45 



for multiplying genera, although they are unable to support them 

 by characters of weight or based on sound principles. 



It is hardly necessary to add that the Mediterranean species, 

 the Livornian at least (as well as the Algerine enumerated by the 

 celebrated Montague in his Cryptogames Algeriennes, \ Annal. 

 des Scienc. Natur.' 2nde ser. torn. x. p. 257, collected unquestion- 

 ably by Roussel), is in all parts conformable to the oceanic spe- 

 cies, of which I have often received splendid specimens from 

 Lenormand, Godey and Auniet, for which reason I shall dispense 

 with recalling their habits and forms. I will only remark that the 

 frond and its divisions are perfectly cylindrical when first taken 

 out of the water, but wither when exposed to the air, and assume 

 a prismatic triangular or quadrilateral figure, the angles of which 

 are very prominent, the sides depressed and channeled. 



At first sight, under the microscope, one would say it was en- 

 tirely composed of round elliptic or oval cells, whether isolated 

 or ranged above one another in parallel rows ; but in vertical as 

 well as horizontal sections, the innumerable filaments which form 

 the central part are easily detected : by taking small slices of the 

 frond in the direction of its greatest diameter and putting them 

 under the microscope, the mode of growth is at once apparent. 

 They thus form a cord, or I might rather say a fascicle, which, 

 like the mealy part of the thallus of some lichens, occupies the 

 centre of the frond, extending even to the furthest divisions. In 

 their course they are repeatedly dichotomous and form two sets 

 of branches, the one directed upwards, the other bent in a hori- 

 zontal direction, so as to unite by their clavate extremities, which 

 are once or twice divergent and bifurcate, with the peripheric 

 stratum of the frond. 



The form of the filaments is rather compressed, the diameter 

 being often unequal and slightly thickened at the commencement 

 of each dichotomy and at their extremities, which inclose a co- 

 loured substance, but are themselves diaphanous and completely 

 colourless. In some of them I have been able to determine the 

 presence of lateral branches of various lengths descending in a 

 winding course towards the inferior part of the frond. I also 

 thought I discovered in the filaments, more especially in the 

 points where they became bifurcate, traces of partitions ; and I 

 can declare, without hesitation, that the superficial cells, from 

 which the walls of the frond spring, originate from the claviform 

 and divergent extremity of the centrifugal branches. I must not 

 omit to remark, that the cells of the peripherical stratum do not 

 all communicate directly with the horizontal threads ; if I am not 

 mistaken, those extremities, in which constrictions frequently oc- 

 cur in the form of articulations, may give rise to new cells, which 



