393 



" oblong, contracted below." Mr. Ball said " clavate, and somewhat 

 contracted below." I called it " subcylindrical." Here we see three 

 different observers using as many different terms. The fact is, the 

 fruits vary considerably in form, even in the same umbel. I have 

 before remarked of the three species, " while immature, the fruits of 

 all are contracted downwards" (Phytol. ii. 15); and Ball expressly states 

 that he describes fruits not fully mature. Apparently this form of 

 immaturity is retained in some of the full-grown fruits of Smithii ; as 

 is certainly the case with those of Lachenalii. The exterior fruits, 

 in the umbellule of Smithii, are rarely of equal thickness from base to 

 summit, when full grown, but not being quite equal throughout, and 

 moreover also somewhat angular, the term "cylindrical" requires 

 the qualifying prefix " sub." In the interior of the umbellule, the 

 fruits are compressed one by another, and (seemingly through their 

 mutual pressure) they assume a more angular and tapering form. 

 Such a fruit is represented in the figure given by Mr. Lees ; and 

 which is, indeed, more clavate than oblong. I presume, however, 

 that the form of the exterior fruits should be considered typical or 

 normal. Apparently this is also the opinion of Koch, who writes of 

 silaifolia " fructibus cylindricis." The exterior fruits are nearly cy- 

 lindrical, the interior being oblong-clavate and angular. 



Having thus gleaned from the paper of Mr. Lees, and applied to 

 scientific use, the little information of value which it can be said to 

 have added to previously recorded knowledge, I next ask myself the 

 question, ' Shall the inaccuracies of the paper be left to that inevitable 

 fate which ultimately befalls all error ? or is it better at once to qua- 

 lify and correct them, in order to check the temporary mischief con- 

 sequent on the propagation of errors ?' There is inconvenience in 

 the latter course. Easy as it is, to state simple facts in concise terms, 

 it is by no means equally easy to apply them in the correction of 

 errors and inaccuracies, without adding quotations and explanations 

 which would greatly extend the necessary amount of text. I will 

 therefore not attempt to quote in detail, and then to correct, all the 

 inaccuracies and untenable statements which occur in the paper of 

 Mr. Lees. Instead of pursuing that course, I will mention some of the 

 circumstances which seem calculated to place the subject of discus- 

 sion in a more correct position ; though even this will unavoidably 

 lead to repetitions and explanations, which have become necessary 

 only through the faults of Mr. Lees' paper. 



