M. Miiller on the Germination o/Isoetes lacustris. 323 



high up with clean sand, the ripe ovules strewed upon this and 

 loosely covered by a thin layer of sand. Germination is sure to 

 follow. If the lower portion of sand be deep enough, the young 

 roots penetrate down with great strength into this layer ; and if 

 the glass be high enough to allow of its being filled up with a 

 considerable depth of water, the young leaflets also grow upward 

 with great rapidity. These deep strata of water and of sand act 

 like a wonderful charm upon the young embryos. If these how- 

 ever cannot attach themselves very firmly by their roots, they 

 come to the surface of the water, and then it is impossible to set 

 them again so as to make them thrive any longer. They are 

 sure to die. 



The reason why I wished so much to observe the formation of 

 the third rootlet was the desire to solve, by indubitable proofs 

 obtained from the course of development, the question once pro- 

 pounded by H. V. Mohl, as to the reason why the youngest roots 

 of the rhizome of Isoetes occur in the centre and the older in 

 the periphery. 



I cannot settle this question from the experience of my em- 

 bryos, yet I believe that it may be brought near to a decision 

 through fig. 24 d. 



There is no question that the root of the third leaflet must be 

 developed in a diff'erent way from the two preceding. These 

 must be formed outwardly, since the leaves which belong to them 

 lie toward the exterior. But the third leaf lies in the interior, 

 between the two first leaflets (fig. 24 d, x) ; consequently its 

 root must also be formed in the interior. Now since the vas- 

 cular bundle of the leaf constantly ran into the root in the two 

 first leaflets, it is probable that the vascular bundle of the third 

 leaf will do the same. Then it must penetrate the rudiment of 

 the young rhizome and make its appearance as the youngest 

 root, between the two older, consequently in the centre of the 

 rhizome. Under these circumstances, we have to wonder less at 

 the peculiar mode of issue of the root than at the phsenomenon 

 that the vascular bundle belonging to every leaf is elongated out 

 from it through the rhizome downward, and there comes to light 

 as a root. The latter consequently has a wholly dependent con- 

 nexion with the leaf. We can therefore as little talk of a prin- 

 cipal root in Isoetes as in the Selaginellce, in spite of Schleiden's 

 opinion. 



A. Braun some time ago* also gave an explanation of the 

 peculiar arrangement of the roots in Isoetes. He said that it 

 might readily be explained if it were assumed that the vessels, in- 



* Flora, 1847, No. 3. 



