1881.] 



MICEOSCOPICAL JOURNAL. 



105 



t 



creting than for absorbing, such as 

 the tomato and Rhodora. 



Tentacular hairs are the most dif- 

 ferentiated and specialized of all, — 

 they being endowed with the power 

 of movement in response to stimula- 

 tion, besides the power of secreting 

 and absorbing. These are best ex- 

 emplified in the Droseracese. 



At the other extreme of the scale 

 of development are what I term pa- 

 pillae, which, together with what may 

 be called villi, are perhaps not so 

 much of the nature of immature or 

 undeveloped hairs, as of organs inter- 

 mediate between hairs and stomata. 

 Instances of the occurrence of these 

 forms — which are hard to separate, 

 the latter being apparently only less 

 differentiated than the former — may 

 be found on many of the Ericacece, 

 particularly on Rhodora Canadensis 

 and Rhododendron Niittallii, and also 

 on Clematis recta and on Porsooni 

 salicina. 



There are, of course, many other 

 forms of hairs, some of them inter- 

 mediate varieties to those I have 

 named, and some of them connecting 

 links between hairs and glands or 

 other epidermal organs ; and several 

 of these often occur upon the same 

 plant or part of the plant. 



It would be an easy matter to 

 arrange a series of specimens show- 

 ing an almost imperceptible running 

 of one of these forms into another, 

 in perfect gradation all the way from 

 the true sunken gland up to the most 

 developed and specialized forms of 

 radiate, stellate, peltate or tentacular 

 hairs. 



The difference between a secreting, 

 glandular hair and a true gland seems 

 to consist mainly in the fact that one 

 is entirely sessile, or sunken, while 

 the other is more or less pedicellate. 

 Sachs states that all reservoirs of oil, 

 resin, gum, etc., whether internal or 

 external, are only slightly different 

 developments of the same organ ; but 

 if we include among such reser- 

 voirs the turpentine-cells of the Coni- 

 ferae, as has been suggested by some 



writers, we shall have to rearrange 

 somewhat our present ideas of what 

 are epidermal organs. 



As may have been inferred from 

 what I have already said, the forms 

 of hairs, glands, etc., are more or less 

 characteristic of families, orders, gen- 

 era, and in some cases of species 

 even. 



This is the most interesting, and 

 scientifically the most important point 

 in this whole subject ; but it has not yet 

 been even approximately elaborated, 

 and there is vast room for research 

 here and great promise of valuable 

 results. 



Take the forms of hairs peculiar to 

 the Elaeagnaceae, for example, they are 

 as characteristic of the order as any 

 other part of the plant ; indeed, in 

 this particular case, I think they are 

 the most characteristic and* distin- 

 guishing part. And yet little or noth- 

 ing is made of these microscopical 

 characteristics in classification, al- 

 though enough has already been 

 learned concerning them to show that 

 they are more intimately connected 

 with the physiological nature and natu- 

 ral affinities of the plant than are many 

 less intrinsic qualities or charac- 

 ters, which are given a prominent 

 place in the usual system of classifi- 

 cation. To be sure, reference is some- 

 times made to the " silvery " appear- 

 ance of some species, and to the " scur- 

 fy " appearance of others, to the naked 

 eye, but the characteristic feature of 

 Shepherdia or Elceagtius or HippophcB 

 is seldom even referred to in such 

 botanical works as I have happened 

 to consult. 



The hairs of the MalvacecB also are 

 very characteristic. In general, to 

 the casual observer, they seem to 

 come under the very inclusive desig- 

 nation of stellate. But they are not 

 stellate, in the proper limitation of 

 that term, or as the hairs of Deutzia, 

 for example, are stellate. In the 

 genus Sida, and in one or two other 

 genera, they are what I have called 

 peltate, and in still other genera they 

 tend toward what I have called the 



