ON THE GEEMIN-ATION OP CHARA. 311 



reverse, or with the broad side nearly at right angles to the plane of a 

 radial section of the leaf which gave it birth In germination there 

 was a corresponding deviation in the position of the first dividing 

 ■wall (halving- wall) being parallel in the majority and at right angles 

 to the broad side in the smaller number. All these observations tend 

 to prove that the position of the wall with regard to the axis of its 

 origin is constant. 



Certainly they do not furnish indisputable proof of this assump- 

 tion, because germination has only been seen in oospores that had 

 fallen away from the place of their origin ; but the centre of gravity 

 of the oospores is so situated that, in failing when ripe, their greater 

 axis lies horizontally on the horizontal flat surface, and their broad 

 face is parallel to it. Occasionally, it is true, they are found in 

 different positions on account of the unevenaess of surface upon 

 which they fall, or from falling against other bodies. Hence arises 

 the question whether the position described really has the morpho- 

 logical relatioj supposed, and is not dependent upon the action of 

 geocentric force. 



Attempts to clear up this point have failed to furnish a decisive 

 result. In the encrusted Charas it happens that the apparently ripe 

 oospores on the leaf bearing them, that is, in the dead and persistent 

 membrane, remain in their original position ; but all endeavours 

 to determine the eventual morphological position of the partitions 

 were fruitless, as they did not germinate. It is, moreover, too difficult 

 a subject to follow up from fallen not flattened oospores, as it is 

 almost impossible to prevent the vessels of water in which they were 

 allowed to drop from being shaken. In the four only examples which 

 remained unmoved — two of C. foetida and two of C. crinita — the 

 halving-wall, so far as could be ascertained, was parallel to the 

 horizontal substratum. 



Respecting the position and direction of the remaining partition- 

 walls of the pro-embryo, not evident from the foregoing descriptions, 

 it may be mentioned that the plane of the halving-wall in the stem- 

 node cuts the first at a wide angle, almost approaching a right angle. 

 An exact determination of the latter is scarcely possible. As to the 

 position of the halving- wall in the root-node, I cannot venture to off'er 

 any opinion in consequence of the prevalent irregularity in the 

 divisions. The first whorl-cell in the stem-node on the initial cell of 

 the first stem, may be either on the side farthest or on that nearest 

 the primary root. Both positions were observed in C. crinita ; still 

 this is a point difficult to clear up with certainty, from the twists and 

 crooks the pro-embryo undergoes in its early stages. 



Concerning the direction of growth in the species examined, the 

 pro-embryo and primary root emerge from the shell and diverge at a 

 very acute angle (fig. 2) ; and only subsequently from the swelling of 

 their points of insertion, especially of the base of the primary root, the 

 divergence increases to 180 degrees (figs. 5, 6). In the Nitellas and 

 Tolypellas examined the first node curves directly after division, and 

 its two primary shoots diverge from the beginning at right angles to 

 the greater axis of the oospore (figs. 12, 13, 18). As soon as the 

 more decided elongation sets in, the familiar geocentric curves of the 

 pro-embiyo upwards and the downward-growing root commence. 



