PHANEROGAMIA : FLOWERS. 433 



often fill berries, it is no longer possible to determine how much of 

 it belongs to the inner wall of the fruit, how much to the conducting 

 cellular tissue and to the funiculus : the whole mass may in any 

 case be termed pulp. 



Upon the more or less distinct production of the layers of the testa, 

 and upon their varied formation, rest all those varieties of the fruit which 

 strike us on a first impression, and which had received their common 

 distinctive names long before botanists had constructed fruit-systems. 

 Where the layers are clearly distinguished, the epidermis of the surface 

 is seldom particularly striking, the inner epithelium frequently shares 

 the transformation of the inner parenchymatous portion, which varies in 

 its consistence from that of leather up to a stony hardness, sufficient to 

 give out sparks when struck by steel ; but in every condition it consists 

 of thickened cells which are usually porous. Two different conditions 

 may be distinguished in the inner parenchymatous layer : 1st, Where 

 several layers of cells go to its formation, the long-diameter of the cells 

 of one layer is commonly at some angle to those of the next layer (as in 

 Leguminosce, Amygdalece, and almost all capsules) ; 2ndly, Where only 

 one layer is present, the cells are so arranged that five, or six, or more 

 cells, lying parallel, form little plates, from which the layer is so con- 

 structed, like mosaic work, that the long-diameter of the cells of one 

 plate is never in a line with that of those of the next plate (as in Ascle- 

 piadacece and Cruciferce). The epithelium of the inner surface is also 

 sometimes changed into elegant spiral-fibrous cells, as in some Papave- 

 racece (Chelidonium\ in Umbelliferce (Anethum\ &c. ; more rarely it is 

 developed into true epidermis with perfect stomates (as in Reseda, Passi- 

 flora, &c.). The outer layer of parenchyma varies from a leathery con- 

 sistence to the most perfect dissolution into easily compressible, succulent 

 cells, DeCandolle and others have taken pains to trace this layer to the 

 texture of the normal leaf. It appears to me that this is empty trifling. 

 In the first place, it has as little of the structure of a normal leaf as of 

 the form ; secondly, many germens do not arise from foliar organs ; 

 and thirdly, in one and the same strictly-defined and quite natural family, 

 the most essential varieties are found to exist in nearly related genera, as 

 in the Solanacece, where true berries and capsules, and in the Dryadece, 

 where true small berries and achaenia occur. 



In the formation of the berry and the pulp, the origin of cells within 

 cells, &c., may usually be beautifully observed. The parent-cells, how- 

 ever, especially towards the time of the ripening of the fruit, become 

 absorbed before the young cells are firmly united together and sufficiently 

 expanded to become firmly connected with the neighbouring cells when 

 they are set free ; hence they remain lying loose in the abundance of juice 

 which is simultaneously produced. A circulation in reticulated, anasto- 

 mosing currents is exhibited in these isolated cells (in Solanacece, Cac- 

 tacece, and Lonicerece). 



Some very thin- walled germens in Aracece and Naiadacece, as also, in 

 part, in those families whose one-seeded, indehiscent germens are closely 

 united with the external integument of the seed, and thus represent what 

 Linnaeus called naked seeds (e.g. Crraminece, Labiates, Boraginacece, Com- 

 positce, &c.), are in the ripe fruit so compressed and so uniformly 

 developed in all their (few) cell-layers, that they can only be classed by 

 means of analogy with one of the types mentioned. 



The epidermis of the fruit in the occasionally dehiscent fruits exhibits 



F F 



