From G. A. W. Arnott 30 June 1848

My dear Sir

I have this morning received yours of the 28^h and have given your syllabus full consideration. Of course being considered a Syllabus of your own course, it does not apply to the method followed in other universities. Thus some mix up the anatomical and external structures - others as yours, separate the two. As for me having both a winter and summer course, I am forced to adopt different plans according to circumstances: Hitherto I have given the internal structure and external form separately - but I sometimes bring in the anat^y of plants immediately after the subject of leaves sometimes postpone it till I have arrived at the ovary fertilisation - and this summer until I had completed my notes of the fruit or maturation and fecundation only. I mention this merely because although your syllabus comes nearest to this plan I myself follow it is not exactly one that would suit me out and out. Otherwise I would without hesitation adopt it as a sort of text book in my class, being unquestionably the best I have seen.

Now with regard to the restricted questions mentioned at p. II of the Introdⁿ, the first relates to questions on technical terms. Perhaps I do not quite understand you: are you to put 200 questions or only to put a few questions on 200 terms? If the former, I think it would be lost time - if the latter I see no objection except that I think the whole might be abridged by throwing down to the Candidate a plant, and desiring him to make it out by some flora: this he cannot do, unless he understands the definitions.

I am opposed in Botany to all mere memory examinations: a man may be adjudged proficient in the study, and not be able to recollect the Diagnoses even of the British Families or small orders. Besides you say under topic 2^d (at p. III) that the cand^e is not to learn them by rote, but to frame them for himself: such at least is what I understand you to say. Now this is not an easy matter, and one that has puzzled learned people that either Hooker, Lindley, yourself or me (if you will allow me to make the quartetto). By what strictly accurate character are we at all times to separate Scrophularineae and Solanaceae? How can a student or Tyro make out the diagnoses, or discover the errors made in our books? I am quite of opinion that several orders are so well marked that they may be defined in a few words, and with these the candidate should make himself tolerably acquainted but where is he get [sic.] these short char^s? or how is he to know what characters may be truly diagnostic, unless our books provide them such. Take for ex. Anonaceae, this order is at once characterised by its distinct petals (or its bisexual flowers) and ruminated albumen. It requires nothing else: some class it with stamens hypogynos and ovaries several superior: we get into a mess, for one shows a superior ovary, and another an inferior fruit. Thus on the other hand it is readily separated from Myristaceae, which never has petals (and is always unisexual) but which has likewise ruminated albumen. I know however plants which have ruminated albumen & bisexual flowers but a gamopetalous corolla.

What has always appeared to me thus of use for a student is to make him acquainted with such short isolated characters as have occurred to myself: and many other botanists might add to them: if we had a tolerable collection of these, and they were published, I would recommend every young botanical student to commit them to memory, and at the same time compare them with specimens or plates so as to impress them on his mind - but I am very doubtful as to the expediency of him bottling up in his brain diagnostical characters of hosts of orders, bound together perhaps by habits and with very empirical distinctions, or at times disjoined by habits but yet having some very marked points of structural resemblance, so as to lead a beginner to confound them.

Perhaps however I quite misunderstand the kind of questions that are likely to be put under topic 2^d. My remarks apply chiefly to such a one as "How is Violaceae to be distinguished" which implies a long character list must be got by memory. If you add "How is it to be distinguished from its allies" - a question arises, what are its allies? Lindley says one thing, another man another; its real allies appear to me (but not all) of Lindley's Violales and among British plants I do not recognise anything that is really allied with it. Still Violaceae (even the regular flowered ones) may I think be distinguished at all times by two or three words, from all other orders, Brit. and foreign. What appears to me to be a better mode of questioning is How is Violaceae distinguished from such or such an order - whatever that be. Suppose Cistaceae. Here the stipules are of no avail - the regularity of the flowers I omit, because some foreign ones have the flowers regular - the single style is of no avail - the number of parts not always of avail - the presence of albumen of no avail, for both possess it in considerable quantity - the placentation not always diagnostic; so that the real distinction becomes very abridged. A question of this kind moreover will imply to me for reflection, and to pass under the minds eye in [illeg.] the general aspect of the various plants. If this be the modus operandi then I think such examinations will bring out rubbish for answers to such can in no way be got by rote. I think however you should add to your selecting of Brit. orders, a few others and also some prominent foreign ones: these as a student gets little benefit from lectures on this till he knows well what nat. orders are, not of England only, but also of Arctic and tropical countries - and although I did prepare some lectures on the subject, I have never used any except the introductory portion - for by the time the student was able to follow me, when I named a natural order, (and unless he does understand me, it is better not to waste the time), I had drawn to the close of the course and found all too [illeg.] on practical botany to pay much attention to general deductions. If I could accomplish it I would devote the whole course to practical botany or what he could not get in books without very extensive reading: and if we had in fact schools of botany this might be done, but they come to me ignorant either of petals, stamens, or pistila.

In passing the examinations ought you not to give a candidate a choice of what branch of botany he wishes to be examined upon. A really good Veg. Physiologist as Treviranus (and sometimes also an anatomist) is very seldom a good botanist, generally not a botanist at all - and if we take Humboldt as a specimen of a Botanical Geographer he has not the smallest pretensions to botany. It appears to me that the candidate cannot with efficiency master all the topics you mention, and the examination may hit on the very one on which he is weakest. If he chooses the anat. or Physo^y I see little occasion to ask him about terms, except such as relate to these points. On the other hand if he selects the systematic department, he may be examined on the distinguishing between a few orders and also on the terms connected with these.

As to your 3^d topic, I see no objection to this, provided it be confined to those who come forward in the systematic department. I doubt however of the utility of the person burdening his memory with the Jussieus subdivisions of Epistamineae peripeteae, monoepigynae &c &c, these being of very little use in practice. De Candolles divisions are of more importance, and ought to be digested well by everyone - only I do not think you have in the table stated De Candolles again quite accurately. He does not divide monocot^s in the Phaner. & Crypto - but he divides Endogenae, into Phaner. & Crypto, that is into Monocot. and Acot.

The simpler divisions are however Acot. - Monocot., Dicot. (both of these are undoubtedly [illeg.]) - and then subdivide Dicot. or Exogen., only as proposed DC. This is in fact the method pursued by you, p. 11-22, in which I quite agree.

After all, this is so egotistical a letter, that I had almost written another but I have not the time so shall let it go.

I see at p. 6 you speak of rhomboidal - dodecahedral cells. Did you ever see any - or in what plant can I find them? I have somewhere read that it is this kind which is so common in pith, and cellular plants, exhibiting (in a section) a hexagon. Now unquestionably a rhomb. dodecahedron does show a hexagonal section, but the cell that is most common is in the annexed, which has 12 four sided faces and two hexagonal ends, in all 14 sides: and this is the one which appears to be generally spoken of as a dodecahedron or Hexagonienchyma. [illustration of 14 sided cell]

I am much puzzled about fusiform cells or Prosienchyma - and have some doubts about this tissue - and I see no reason why those whose apices touch each other without overlapping differ from ordinary Parenchyma except by being a little sharper, while those which locally overlap are Plurenchyma. When the first kind have flat faces they may easily be confounded with Plurenchyma because cells 1, 3 may be said to overlap but I think incorrectly - truly overlapping cells (and with Plurenchyma), I have never met with. [2 diagrams of cells]

Yours faithfully

G. Walker Arnott

I prefer calling ducts by the name of Angienchyma and [illeg.]