ORIGIN OF ENDOGENS PROM EXOGENS. 525 



ences of an aquatic medium. It is only by taking a large and 

 comparative survey of endogens that one can come to any general 

 conclusion on the subject. 



First, then, let us observe that of the truly aquatic orders, while 

 some, such as HydrocJiaridece, Butomea, and Alismacece, have 

 conspicuous flowers, a large number of genera included in 

 Juncaginece, Potamece, Naiadece, Typhacece, Juncece, Cypcracece, 

 Restiacece, &c. are very greatly degraded. 



Hence, the evidence from coincidences is accumulative, w T hile 



comparative anatomy justifies the conclusion that there has been 



a distinct " cause and effect " in the reduction of the floral 

 structures. 



In looking for general peculiarities among the inflorescences 

 of endogens, the spadix or dense, mostly unisexual spike is 

 observable. Thus, e. g., it occurs in Aroidece, Potamogeton, 

 Typhacece, Pandanus, Cyclanihece, Aspidistrece, &c. In those 

 cases the perianth is generally greatly reduced or wanting, as in 

 Arum maculatum, while the sexes may be separated as in that 



species. 



Turning to Alismacece, Ac, with their more highly developed 

 perianth, we seem to see the origin of some of the spicate forms ; 

 for Alisma Plantago, Sagiftaria sagittcefolia, &c. supply whorls. 

 Juncaginacece, to which Aroidece are allied, have racemose spi- 

 rally arranged flowers. Supposing these to be reduced to com- 

 pact spikes by the arrest of the pedicels, we arive at the spadix, 

 the perianth becoming arrested, in consequence of the flowers 

 being densely compacted. This, it will be remembered, is a very 

 general result elsewhere, when flowers are very crowded, as far 

 as the calyx is concerned. Thus it is more or less aborted in 

 Umbelliferce and all the Gamopetalce with inferior ovaries, as well 



as Rhododendron, Syringa, &c. 



The fact that Sagittaria has male flowers above and w r horls of 

 females below, is suggestive of the origin of a similar arrangement 

 as occurs in Arum maculatum. 



The arrest of pedicels and the conversion of panicles to spikes 

 with various degrees of reduction of the perianth to bristles and 

 hairs, as occurs in Typhacece and Cyperacece, &c, or to scales 

 (lodicules *) in Graminece, are gradations correlated on the oue 



With regard to any use these diminutive organs may have, I would 

 suggest that they are little reservoirs of water. They always contain it, and 

 sometimes have bladdery aqueous cells on their surface. 



