588 



in one specimen of A. littoralis I have found a few horizontal seeds, 

 which were probably produced by hermaphrodite florets : such a 

 case however is exceedingly rare. M. Moquin-Tandon says that 

 the sepals or, as he calls them, bracts of the fertile flowers are altoge- 

 ther separate. I do not find this to be the case : they still remain 

 slightly united at the base. Mr. Babington supposes A. nitens to 

 have been introduced accidentally at Ryde in the Isle of Wight, the 

 only British habitat. 1 do not know if it has ever been cultivated as 

 a spinach. This has doubtless been the case with the A. hortensis, 

 another species of this division which has not been naturalized in 

 England, nor perhaps in Europe. It has not the silvery scales on 

 the leaves which characterize the preceding species. The third spe- 

 cies of this division found in Europe is the A. vincta, said to be cul- 

 tivated about Verona, but according to Willdenow is wild at Venice 

 on the shores of the Adriatic ; adding however that the native plant 

 differs "toto cgelo" from the cultivated one. Willdenow distinguishes 

 this from the two preceding by the toothed sepals of the fruit; Mo- 

 quin-Tandon, from hortensis by the leaves glaucous beneath, and the 

 lanceolate form of the upper ones, and from nitens by the even, not 

 reticulate, sepals. M. Moquin-Tandon puts the Veronese plant 

 among those not sufficiently known, and does not notice the Vene- 

 tian one. 



The second addition is marina. Linnaeus in his Mant. 11, gives a 

 plant under this name as found in England, and distinguished from 

 littoralis by its serrated leaves and small stature. Babington makes 

 no mention of the latter peculiarity, but adds that the perigone is 

 closed, while in A. littoralis the sepals are spreading. I do not how- 

 ever find that the sepals of A. littoralis are usually spreading, except 

 that they are often a little turned out at the top. It is acknowledged 

 that the serratures of the leaves form not an absolute, but only a 

 comparative character, and the species must be considered as doubt- 

 ful. I find only A. littoralis in the neighbourhood of Lewes; but. 

 marina seems to be plentiful on the banks of the Thames. Moquin- 

 Taudon doubts whether littoralis be anything but a variety of A. has- 

 tata ; a conjecture for which I see no grounds whatever. 



The preceding species offer no peculiar difficulties. Those with no 

 hermaphrodite flowers, and a tendency to produce hastate or triangular 

 leaves, are not so easily arranged. There are some particulars in the 

 fructification which ought to be understood before we proceed to 

 their investigation. In the first place, the size of the seeds is not 

 uniform, and in some species there are distinctly two forms. Those 



