EL]&MENTS DE BOTANIQUE. 457 



and " Stigmatees," is to insist unnecessarily on an obvious but 

 minor point of detail ; and a similar criticism will apply to a 

 character by aid of which the Angiosperms are separated into three 

 classes, even though that character should prove to be general. 

 The three classes are entitled " Liorhizes Monocotylees ou Mono- 

 cotyledones," " Liorhizes Dicotylees," and " Climacorhizes ou 

 Dicotyledones." That is to say, the primary character is taken 

 from the manner in which the young root loses its root-cap, whether 

 the latter comes off clean (Liorhizes), leaving the functions of 

 piliferous layer to be performed by the outer layer of the cortex, as 

 in Monocotyledons and most NymphaacecE, or whether, as in the 

 rest of the Dicotyledons, the most internal layer of the epidermis 

 remains to produce the root-hairs after the exfoliation of the outer 

 layers as the root-cap. The term "climacorhizes" refers to the 

 step-hke appearance at the root-tip, resulting from the peeling off 

 of a gradually lessening number of layers of root -cap from below 

 upwards, and will be understood on reference to a figure of the 

 growing-point. The result is to separate the grasses from the 

 Monocotyledons, and the NymplmaceijB from the Dicotyledons, and 

 unite the two families into an intermediate class. The second 

 cotyledon of the grasses is found in the projection which sometimes 

 occurs on the axis on the outer side of the embryo. Such an 

 arrangement might provide excellent material for discussion at an 

 evening meeting of a scientific club or society, but is distinctly out 

 of place in a comparatively elementary text-book. Graminece form 

 a well-defined and natural family, which is perhaps less_ closely 

 allied to the CypemcecB than a student would be led to imagine 

 from their close approximation in a systematic arrangement, but to 

 remove them completely from Monocotyledons and to unite them 

 with the polypetalous acyclic-flowered water-lilies is to ignore floral 

 and other characters which have been held to be among the highest 

 in value. The author shows a similar want of appreciation of 

 relative value of characters in making four orders of Monocotyle- 

 dons and placing JuncacecB, Lillaceai, and Amaryllidacem in separate 

 orders. Similar criticism will apply to the subdivision of Dicoty- 

 ledons, which depends entirely on degrees of differentiation of the 

 ovule. Five orders, including a large number of small families, the 

 names of many of which will sound strange to English teachers and 

 students, but among which will be found Lormithacece, SantalacecBf 

 and Balanophorea, are separated as Sub-class I. — " Inseminees." 

 The great majority are considered to have seeds, and form Sub- 

 class II. — " Seminees " — which fall into two orders, according to 

 the presence of a single or double integument in the ovule. The 

 result is that while the smaller groups are generally fairly natural 

 in themselves, nearly allied families are often widely separated, as, 

 for instance, Castanacem (including oak and sweet chestnut) from 

 Corylacem and Betidacem. It is as if an attempt had been made to 

 arrange the different famiUes in a scheme of three dimensions, 

 which by a process of flattening and reduction to a linear series 

 has become egregiously distorted. ^^ ^ Rendle. 



Journal of Botany.— Vol. 36. [Nov. 1898.] 2 i 



