ZOOLOGY A.ND BOTANY, MICROSCOPY, ETC. 991 



being seven ; they do not appear to I e formed in any regular order, and 

 have no relationship to the two carpels. Only a small proportion of 

 them develops into embryo-sacs capable of im^jregnation ; in the speci- 

 mens examined which were parasitic on Poj)ulus laurifoUa and ^sculus 

 Pavia, the number of embryos and embryo-sacs was always two or 

 three, but it is stated to vary according to the species of the host. 



The embryo-sac always consists in its early stage of two cells result- 

 ing from the transverse septation of a mother-cell ; and, as in most other 

 Phanerogams, the lower of these two cells alone developes into the 

 embryo-sac, putting out a lateral protrusion which penetrates into the 

 parenchyma of the ovary, and developes into the broad lower end of the 

 sac. The upper of the two sister-cells does not, however, entirely dis- 

 apjjear, as in most other flowering plants, and may possibly also 

 sometimes develope into a fertile embryo-sac. This confirms Goebel's 

 view that every daughter-cell of an archespore may potentially develope 

 into an embryo-sac. In the mature embryo-sac are seen three antipodals 

 with thick membranes, three bodies belonging to the egg-apparatus, and 

 two central nuclei which afterwards unite into one. 



The above description closely corresponds to that of Treub respecting 

 Viscum articulatum, excepting that in the latter case the upper of the 

 two sister-cells disappears altogether. In both species the embryo- 

 sacs are developed from the hypodermal layer of cells of the end of the 

 floral axis. 



The male flowers of the mistletoe are much less common than the 

 female. Here also it is the hypodermal layer of the anther from which 

 the pollen-cells originate. In the course of their formation the epi- 

 dermis undergoes irregular periclinal divisions, and ceases to exist as a 

 special layer. The outermost layer of the archespore-cells developes 

 into the tapetal cells, the inner layers undergo further increase of size, 

 and become the mother-cells of the pollen. The mode of formation of 

 the pollen in the mother-cells presents nothing special. 



The great peculiarity of the male flowers of Viscum, viz. the direct 

 formation of the anther on the perianth, and not in connection with a 

 special staminal leaf or stamen, apj)ears to belong exclusively to the 

 Loranthacese ; but, within this order, occurs also in Arceuthohium, and 

 apparently also in the very rare Castrsea falcata. 



Arceuthohium.* — Mr. T. Johnson has carefully followed out the 

 embryogeny of this genus of Loranthaceae, especially of A. Oxijcedri. 

 The following is a summary of the results. 



There is formed in the ovary, at the time of pollination, a conical 

 papilla projecting free from its base, containing two embryo-sacs im- 

 bedded at the side of the apex, in which the contents are arranged as in 

 a normal angiosperm. The embryo-sacs arise in each case from a single 

 hypodermal archespore-cell. The morphological value of the contents 

 of the ovary is the same as in Loranthus spliserocarjMS, as described by 

 Treub,| the papilla consisting of the modified apex of the floral axis, and 

 constituting a placenta bearing two buried ovules reduced to embryo- 

 sacs. At no time does the papilla fuse with the wall of the ovary ; its 

 apical region becomes a pseudo-calyptra to the solitary embryo, which is 

 straight, and has an exserted radicle without a root-cap. The dehiscence 

 of the fruit is finally due to the rupture of a basal horizontal merismatic 



* Ann. of Bot., ii. (188S) pp. 137-60 (1 pi.). f See this Journal, 1882, p. 363. 



3x2 



