by the aid of Polarized Liyht. 267 



The lougitutlinal section of the cell-wall behaves to polarized 

 light exactly like the cross-section ; hence, on the application 

 of the plate of selenite, the same colours make their appear- 

 ance, according to its position. J3ut it is far less easy to ob- 

 serve this phienomenon in its purity here, than in examination 

 of cross-sections, because the preparation of delicate longitu- 

 dinal sections of uniform thickness, especially of ])roscnchyma- 

 tous tissues, is far more difficult than the cutting of good 

 cross-sections. Too thick a slice produces colour even by itself, 

 whereby the colours produced by the plate of selenite arc ren- 

 dered more or less impure. A further disturbance frequently 

 arises out of the circumstance that the longitudinal section often 

 takes the side-walls of the cells in an oblique direction, the sec- 

 tions curve round, &c., from which irregularities the colours 

 not rarely undergo alteration, even into the complementary 

 colours. All this renders the examination more difficult; but 

 when good preparations are examined, the above rule will be 

 found contirmed. 



Those side-walls of cells which are viewed from the surface, 

 and arc traversed by the polarized light in a perpendicular direc- 

 tion, exhibit extremely diverse aspects according to the variations 

 of their structure. To observe these phsenomena in pei-fect 

 purity, it is best to select such elementary organs as have their 

 secondary membranes split into fibres, — best of all, spiral ves- 

 sels. Among these, the spiral vessels of the scape of Musa pa- 

 radisiaca afford an unsurpassable material, since they are easy to 

 isolate, of considerable size, and regular structure. ^Vhen one 

 of these vessels, its spirals being drawn somewhat apart, is placed 

 with its long axis in a position perpendicular to one of the 

 Nicols, the fibres on the upper side of the vessel ascend towards 

 the left, those of the under side towards the right, and when the 

 sclenite-])late is interposed, they exhibit the complementary 

 colours. If the colour of one of the strata of fibres is compared 

 with the colour of a section of a cell-wall which is placed in the 

 same direction with the fibres, an agreement in colour is dis- 

 played. If the vessel is rotated horizontally so far that the fibres 

 on one wall come to lie perpendicular to one of the Nicols, the 

 colour of these fibres vanishes so far as they lie in the said 

 direction. It is clear, therefore, that in these fibres one nca- 

 tive axis lies parallel with their longitudinal extension, the other 

 perpendicular to this. It is then very easy to explain why spiral 

 vessels which are inclined with their longitudinal axes at an angle 

 of 45° to one of the Nicols, exhibit a totally different aspect, 

 according as their fibre describes a more or less stcej)ly-ascending 

 sj)iral. Three cases may be distinguished. When the spiral- 

 fibre describes a very slightly-ascending spiral, when therefore 



