376 ON CHORISIS. 



synpetalous, and if you please, synandrous, syncarpellous, whilst sepa- 

 ration may be expressed by aposepalous, apopetalous, (J-c. Adherence 

 arises from pressure of the circles on each other, or expansion of the 

 torus or receptacle, so as to adhere sometimes outward on the lower 

 part of the calyx, sometimes inward on the combined carpels, some- 

 times in connection with both, so as to place the fruit below the other 

 circles of the flower and produce the epigynose structure — it readilj 

 explains many phenomena of common occurrence in flowers. 



Regularity and irregularity of flowers depend entirely on the equal 

 or unequal distribution of nutriment to the parts of the successive 

 circles, the causes of which dififerences are often undiscernible, though 

 the fact is certain- Sometimes the more developed parts are in all the 

 circles on the same side of the flower ; in other cases the opposite sides 

 aye enlarged alternately. In other instances theirregularity is produced 

 by an opposite pair being enlarged in each circle (where the whole 

 number of parts is evenj, or by this arrangement being alternated in 

 the successive circles. It must be evident how many modifications of 

 flowers are explained by these considerations. 



The primary law respecting number is found in the tendency to the 

 number three in the circles of mouo-coty ledonous plants, and to five in 

 those of dicotyledonous plants. The first is an ultimate law of the 

 organization of plants abundantly established by fact, but hardly capa- 

 ble of being connected, so far as we can at present see, with anything 

 else we know'of their nature. It may be doubted whether the second 

 is not connected with the first in as much as one cotyledon or primor 

 dial leaf is found to imply a circle of three parts, two would therefore 

 be expected to produce six, but this supposes the combination into 

 one of two circles of three. Now we have other examples of this sort 

 of combination of circles of parts exhibited to us by certain anomalous 

 flowers, in sufficient number and variety of cases to suggest a sort of 

 rule as to what is likely to happen, and from them we infer that in 

 ordinary cases one part would be lost in the union. That under con. 

 siderable pressure a part would be lost at each point of junction or two 

 in the combined circle, whilst very close position, with circumstances 

 unfavourable to development, such as give us occasional examples of 

 two and one part in a monocotyledonous plant might occasion any of 

 the lower numbers to occur in a dicotyledon. I found the explanation 

 here given of the prevailing number of dicotyledonous plants on the 

 careful examination of a considerable number of those monstrosities, not 



