PIIANEROGAMIA : FLOWERS. 415 



quite imperfect. They show that in those parts of the pollen-tubes which 

 have penetrated into the cylindrical cavity, a process of cell-formation 

 soon begins, so that four cells are formed, which, parallel to the pollen- 

 tube and to each other, extend in a cylindrical form ; then in the free 

 extremity of each another cell is formed (Juniperus communis}, which 

 soon developes into three (?) cells (Abies excelsa\ so that the embryonal 

 globule now consists of twelve cells placed side by side in four rows. The 

 process of multiplication of the cells then advances, and a little papillose, 

 cellular body is formed as embryonal globule, which is seated on a long 

 suspensor composed of four parallel cells. The cells of the latter continue 

 for a long time to expand exclusively in length, and thus acquire a tor- 

 tuous condition in the too short cylindrical cavity. At the place where 

 they come forth from the large cells (corpuscula\ some cells also appear 

 to be formed sometimes, or the neighbouring cells compress the cavity of 

 the pollen-tube together ; in short, there is very soon no further trace of 

 the originally free communication to be discovered. 



Special deviations besides so mentioned I have not yet met with, nor is 

 it probable that differences should occur in the essential particulars when 

 it is remembered that the peculiarities which distinguish the Cryptogamia, 

 Rhizocarpece, and Phanerogamia are far greater than the main points 

 which appear to occur throughout the whole animal kingdom ; yet the 

 Phanerogamia agree so closely in all the rest of their organisation, that 

 it is very improbable that they should exhibit important modifications in 

 so essential a point. In the Plate IV. of this volume I have given a series 

 of the most instructive, and easiest to repeat, observations ; especially in 

 Epilobium angustifoUnm, Orchis latifolia and Morio, Martynia cHandra, 

 Salvia bico/or, (Enothera rhizocarpa, acaulis, and Momordica Elate- 

 rium. I will only add a few words on the preparation of such dissections. 

 If the seed-buds are not very closely enveloped, and immoveable in the 

 germen, I extract them, take them between the thumb and fore-finger in 

 such a manner that with a sharp razor I can cut them accurately in half. 

 To obtain these halves symmetrical, and that the section may hit the 

 micropyle canal, or come near enough to it, I place the seed-bud in the 

 proper position between the fingers, if necessary, with a lens. The two 

 halves thus obtained I place one behind the other, the cut surfaces 

 toward the thumb, between the thumb and finger again, and cut with the 

 razor the thinnest possible slice from the surfaces of the sections. I then 

 bring these two slices under the simple microscope, and dissect the parts 

 with fine needles and knives, if they are not already made evident by the 

 section, which is of course always the best. In one-seeded germens the 

 same may be done if they are very small ; in other cases I at once cut as 

 fine slices as possible, e. g., in the Gourd. 



It is evident that one must always previously study the structure of 

 the unimpregnated seed-bud and germen, and the form of the pollen -tube, 

 and by careful observation make oneself acquainted with the periods of 

 fecundation. Patience and perseverance will always have to be applied as 

 the most important of all means of success. A hundred such slices as I 

 have described may often be made and nothing seen, and perhaps the 

 hundred and first will be so good as at once to complete the investigation. 

 I do not think the method of dissecting off the parts of the seed-bud from 

 without inward, under the simple microscope, advantageous, because 

 much more is destroyed, and in particular displaced, in this than in one 

 simple section made with a sharp instrument. 



