INTRODUCTION TO CRYPTOGAMIC BOTANY. 555 



South of Europe and North Africa, Oregon, Madras, New 

 Holland, and Brazil, possess representatives of the genus. M. 

 polycarpa is remarkable for having numerous receptacles half- 

 way up the leaf-stalk. The species are of no ascertained utility. 

 The fossil Sphenophyllum probably belongs to Equisetacew* 



635. Salvinia agrees with the two former genera in the 

 character of the rhizoma and radical fruit. The fronds, how- 

 ever, have a fern-like aspect, but they are remarkable for 

 having warts radiating from the costa, and crowned with a 

 little tuft of jointed threads. The veins are visible below, and 

 are reticulated. The fruit consists of globose fluted sacs, con- 

 taining a little central columella, to which are attached, in 

 separate sacs, the antheridia and sporangia. The antheridia 

 are globose and pedunculate, resembling strongly those of a 

 Jitngermannia; the cellular-walled sporangia are monosporous. 

 The spores germinate after the fashion of Marsilea and Filu- 

 laria ; but the antheridia are like those of Azolla. The stem 

 has cavities, as in those genera, with a central bundle of ves- 

 sels. It is probable that all the supposed species are reduci- 

 ble to one, and if so it occurs in the warmer parts of Europe, 

 and in the tropics of Asia and South America. At any rate, 8. 

 natans is found on the Niger, in the East Indies, Brazil, 

 and Quito. 



636. Azolla agrees with Salvinia, to a certain extent, in 

 the antheridia, but has the habit of a floating Jungermannia. 

 The antheridia are contained in a common sac, as in Salvinia, 

 and have a similar form to those of that genus ; but they con- 

 tain four or more masses, analogous to the pollen masses of 

 Asclepiads, fringed with barbed pellucid threads (Fig. ] 23), 

 first, I believe, observed by Mettenius, or furnished on one 

 side with irregular root-shaped appendages. The sporangia 

 hang down from the frond, consisting, at their base, of a vari- 

 able number of lobes or appendages. Beyond this is a sort 

 of ring, and then the naked spore connected by a villous 

 thread with the parent plant. The spore in itself resembles 

 those described above, and, probably, germinates in the same 

 manner. This point, however, is at present obscure. The 



* See Ettingshausen Steinkohlen, Flora von Radnitz. Wien, 1855. 



