488 SYSTEMATIC BOTANY. 



laxus intm spiroideis vasis pertensus). The view that these cellular layers 

 are formed before the nucleus requires confirmation, and probably only 

 depends upon the observation not being commenced sufficiently early. In 

 the upper persistent portion of the ovule, or the nuclear tubercle (amnios 

 R. Br , nuclear tubercle Schleid.), Miquel has found two or more embryo- 

 blastic vesicles (cavitates Miq., corpuscula R. Br.) arranged around the axis of 

 the organ, passing unnoticed Schleiden's description of the Coniferous ovule, 

 according to which these originate in the upper portion of the endosperm. 

 He expressly states that the nuclear tubercle, or rather the embryoblastic 

 vesicles, correspond to the embryo-sac, and not to the cavity in which the 

 albumen is formed, which, according to his view, would therefore be albumen 

 formed within the nucleus. He regards the nuclear tubercle as " a com- 

 pound amnios" the "individual embryo-sacs" of which would be the embryo- 

 blastic vesicles. Moreover, Miquel has shown in the seeds the con- 

 nexion of the coiled funiculus with the embryoblastic vesicles, as also the 

 anastomoses of the former; and lastly, has found that the form of the 

 embryo is different in all the four genera, by which circumstance they may 

 be distinguished. In the supplement (iv, p. 79), although he has never 

 been able to discover the embryoblastic vesicles in the unimpregnated ovule, 

 he admits the truth of R. Brown's observations (Ann. Nat. Hist., 184-4, 

 May), according to which it may be formed independently of impregnation. 

 This observation is confirmed by the corresponding statements of Gottsche 

 (Bot. Zeit. 1845, p. 402), who has written a detailed critical memoir upon the 

 inflorescence of the Cycadese and the Coniferse (ibid. Nos. 22-27), in which 

 his observations were extended to living Cycadese. According to Gottsche, 

 the embryoblastic vesicles, which in Cupressus are merely simple enlarged 

 cells of the endosperm, in Macrozamia and Encephalartos, in which they are 

 I'" in length and about \'" in breadth (pp. 399-400), possess a cellular wall, 

 which, however, may constitute a later stage of development. Gottsche has 

 not, however, succeeded in bringing into accordance the contradictory views 

 upon the function of these sacs in the act of impregnation by new observa- 

 tions ; although, in opposition to Hartig and in favour of Schleiden, he 

 believes (p. 417) that the pollen-tubes penetrate the embryoblastic vesicles 

 in the Cycadese also. Although I have no hesitation in admitting that this 

 point has been satisfactorily determined by observation in the Coniferae, yet 

 there is a wide interval, which has not yet been filled up, before we can 

 arrive at Schleiden's view, that the pollen-tube is prolonged inwards as 

 far as the embryo-blaston itself, with which neither the illustrated de- 

 scriptions of Brown or Miquel of the origin of the embryo-blaston, from a 

 spherical cell inclosed in its vesicles, and resembling a pollen-granule, can 

 be brought into connexion. These two figures, one of which is taken from 

 the Coniferse, the other from the Cycadese., agree so completely that they 

 cannot be doubted. In my opinion they only admit of one interpretation, 



