52G EEV. GEORGE HENSLOW ON A THEORETICAL 



hand with, and may be partly traceable to an aquatic medium, 

 partly to the absence of insect visitors, and partly to the crowding 

 of the flowers. 



Lastly, a feature correlated with the whorled inflorescence of 

 Sagittaria, &c, is the fact that in many cases the dense mass of 

 flowers on spadices is not arranged spirally, but in alternate 

 whorls. This is the case, e. y., in Potamogeton natans, the whorls 

 being in threes, and fourteens in Acorus Calamus *. 



Another peculiarity of several genera of endogens resides in 

 the quaternary arrangement of the floral organs. Thus, it is 

 found in Tetroncium, Potamogeton, Caulinia, Naias, Carhidovica, 

 GymnostacJiys, Aspidistra, &c. The question arises, how has the 

 number M four " arisen ? In the work mentioned I offered the 

 explanation of quaternary flowers in exogens as generally result- 

 ing from an opposite and decussate condition of the leaves ; and 

 I think the same cause will apply to endogens, though the leaves 

 of endogens are never, as far as I know, opposite and decussate, 

 excepting the first pair, as M. Ph. van Tieghem has shown to be 

 the case in Asparagus. 



Before I return to the preceding suggestion, it will be as well 

 to show why other resources for a quaternary type must be 

 rejected. The first is * symmetrical reduction "f. Though fives 

 are typical of, say, Elder-flowers, sixes and fours may be often 

 met with as the results of symmetrical increase and reduction 

 respectively. Similarly, sevens and fives are to be found in 

 Lythrum Salicaria, in which sixes are usual or normal. Now in 

 endogens threes are normal. Therefore fours might be expected 

 to be found occasionally by symmetrical increase ; but there is 

 no ground for supposing fours could thus become established 

 in so many genera of different orders. 



Turning to phyllotaxis, the same or " primary M J series, repre- 

 sented by the fractions |, £, -f , -|, &c, applies to endogens as 

 well as to exogens, the leaves being usually either distichous or 

 tristichous ; but fours are not obtainable as cycles from any one 

 in this series. If the secondary series ^, ^, ± &c. ever occurred 

 in endogens, then " fours " might have resulted, just as " threes M 



* Similar differences occur among the cones of Cycadea. 

 t "Note on the Causes of the Numerical Increase of Parts of Plants." 

 Journ. Linn. Soc. 1877. 



% " On Variations in the Phyllotaxis of Kclia?ithus tuberosus? Trans. Linn. 

 Soc. vol. xx vi. p. 647 (1868). 



