as distinguished from the Symplocacese. 137 



but at the same time I shall be able to show that m Halesia the 

 remarkable development of its seed results from an ovary exhi- 

 biting exactly the same normal structure as that of Styrax, and 

 that it offers no analogy whatever with the fruit and seed of the 

 Symplocacece. 



The structure and growth of the ovary and fruit of Halesia 

 have been misunderstood equally by Gaertner, Don, Endlicher, 

 and DeCandolle. In H. tetraptera the upper moiety of the ovary 

 is quite free, rising in a conical form above the mouth of the 

 calyx, and is completely unilocular within ; inside its lower or 

 adnate moiety, at its base, we find a depressed placentary axis 

 connected with the walls of this portion by four very short, thin 

 partitions, so that it is here spuriously four-locellate ; and upon 

 the short placentary axis, in each of these divisions, are seen four 

 oblong ovules, two of them superior and standing erect, the 

 others being pendulous within the cavity, all being attached by 

 the short and shtirp point of one of their extremities : laterally 

 these ovules are separated by an interval; but vertically the 

 points of attachment of each upper and lower ovule are approxi- 

 mated upon a minute prominence of the placenta. The space 

 above the placenta is completely unilocular, comprising five- 

 sixths of the entire length of the ovary ; and the four short basal 

 partitions are seen continuous with as many parietal nervures, 

 that extend thence to the summit, where they are prolonged for 

 some distance into the hollow style. This structure, irrespective 

 of the number of ovules and partitions, agrees precisely with 

 that of Styrax, with this difference, that in the latter genus the 

 ovules are inserted upon separate tuberculiform processes ema- 

 nating from the placenta, while in Halesia they are wholly free 

 and attached by a point. 



The growth of the ovary in Styrax and Strigiliaj as before 

 mentioned, is coniSned entirely to the upper portion, while the 

 lower moiety remains of its original length. In Halesia, the re- 

 verse is observed ; for at the same time that the upper moiety 

 continues quite stationary, appearing afterwards like the swollen 

 base of the style, the lower half acquires a gradual increment, 

 until, by the period the fruit is perfected, it attains at least 

 twenty times its former length and breadth. The changes that 

 occur during this growth may be noticed distinctly about a 

 month after the fall of the corolla, by which time the ovary has 

 acquired double its former proportions. On making a section 

 of the ovary at this period, it will be found that, owing to the 

 much greater increment of its basal portion, the placenta, instead 

 of appearing to originate, as formerly, from the very bottom, is 

 now elevated considerably above the base of the central space, 

 and onC; two, or three of the incomplete cells are seen raised up 



