140 INTRODUCTION TO CRYPTOGAMIC BOTANY. 



peculiarity which Thuret observes exists in many Algae, which 

 frequent running streams, the motile bodies being thus enabled 

 to contend against the current. These fragmentary threads 

 divide longitudinally and transversely, at last constituting a 

 bundle of new threads, which gradually by increase of the 

 gelatinous or filamentous elements, assume the normal form of 

 the species. It is to be observed that, in an early stage of 

 growth, each thread has its own coat of gelatine. The genus 

 Hormosiphon (Fig. 21) is therefore not separated from 

 Nostoc with sufficient reason. Derbes has recorded the trans- 

 formation of the moniliform joints into zoospores. This, how- 

 ever, requires confirmation, and it can scarcely be doubted that 

 the fruit, whether consisting of zoospores or not, must be looked 

 for in the enlarged joints. The mode of increase observed 

 by Thuret is probably not strictly propagation, but is rather of 

 the same nature as that in Palmellece. 



1 13. In Nostoc there are, as just said, individual joints which 

 exceed the rest in diameter. These larger joints will in all 

 probability eventually prove to be connected with the fructi- 

 fication. But these swollen joints are still more remarkable in 

 Anabaina and its allies, where they are also accompanied by 

 connecting cells, differing from the others in size, but not 

 acquiring a dense endochrome. In SpJicerozyga Car'michaelii 

 the fruit-cells are oblong, and their endochrome of a vivid 

 green, in S. Thwaitesii, Broomei, and Berkeleyana, their form 

 is more elliptic, and their colour brown. They are sometimes 

 solitary, but more frequently they follow each other in conti- 

 nuous chains, or in abrupt groups, consisting of from two to 

 five individuals. The connecting cells are sometimes clothed 

 with cilia, in which case they may be either terminal or 

 situated indifferently. In the last-mentioned species the 

 young threads are inclosed, two or^three together, in a mucous 

 sheath. This prepares the way for Spermoseira, in which 

 there is constantly a membranous tube to each thread. The 

 joints are short, and less decidedly moniliform, the fruit cells 

 more inclined to be globose, and the connecting cells some- 

 times, though not constantly, pale rose coloured. They ap- 

 proach very near to Oscillatorece. 



