The inner epidermis of the onion bulb cataphylls

(the onion skin)

Easy and not so easy methods to work with

Walter  Dioni              -              Cancún, México

6) – fixing with Clarke’s fixative -  Staining with Blue 1, and Eosin

Follows from part 5 – Fixing with Alcohols


Fixatives formulae

We have reviewed (in the first 5 articles) some physical factors and some chemicals, used alone, as fixatives. Of course, many more products were tested, by hundreds of histologists, through decades of histological practice. It is clear that the temptation to mix, and try to combine the characteristics of the different reagents, to improve the fixation of the investigated tissues, must have occurred very early, if I judge by my own tendency to try just now.

The best reagents needed for success were selected over a long period. Even now new reagents are incorporated to meet the requirements of new targets. And many others had been discarded under the pressure of novelties, availability, prices, or, as in the last decades, social considerations.

I think it might be instructive to review some of the earlier attempts, to understand the efforts of our earlier predecessors.

And we can make these, with little effort, with the reagents at the reach of an amateur.

Allow me a small recapitulation: I have tested boiling water, and 60ºC water, as fixatives for the epidermis of the onion cataphyll (first article), later I followed the “Ceviche’s track”, assessing citric acid (2nd article), and, after this, the well known acetic acid.(3rd article) I accepted the 60ºC water as a good fixative, and the citric acid as a good approximation. But I rejected the boiling water, and only partially accepted the acetic acid (as a nuclear fixative), used alone in the rough conditions of these tests.

96% Alcohol (4th article) has shown a capacity to fix and make evident the cytoplasm, better than the acids do.

I also tested, and approved, at least for my use, the Blue 1 food dye, as a dye for the onion cells, because it gives fairly good images of the cytoplasm and very good pictures of the nuclei. (May 2011 Micscape )

At first sight (after the tests I have presented formerly) what should seem disappointing, because it stands for a long time as an histologists favourite ingredient, is the fact that alcohol, especially 96% alcohol, as shown in the previous article, has given a better representation of the cell as a whole, than acetic acid, even weakly shows.

After so many years of use by so many histologists, there is a long time that a consensus was attained, and is now totally accepted that the acetic acid, used alone, is not a good general fixative.

It seems that it fixes cytoplasm very badly, especially in animal samples; it swells cells, and dissolves totally or partially, some components, even RNA (which includes not only the nucleoli, but also many cytoplasmic granules and the mitochondria), and you can only use and recommend it as a mitotic nucleus fixative, a fixative for DNA. (This is why the derogatory affirmation of Bolles Lee, did you remember? If not... read the former articles.)

But, in itself, this specialization is good, and this is why (as they will be seen by my two readers, whether they followed me and reviewed the formulae as I recommended), acetic is part of most of the best histological fixatives of general use.

By the way, without to go any further: one of the histological fixatives preferred by modern botanists for general use (when they discard formaldehyde) is what many call Carnoy 1, and which others (also many), who respect the chronology and priorities, and especially the zoologists and human histologists, call Clarke’s Fixative (1851). The year in brackets indicates the date on which the author published the formula. This is the citation of the original publication:

CLARKE, J. L., 1851. 'Researches into the structure of the spinal cord.' Phil. Trans., 141:607

Carnoy, used and published in 1886 (35 years later!), the formula of the Clarke’s (alcohol plus acetic acid in a 3:1 ratio), reinventing it, or, most probably, copying it without reporting the origin), and in 1887 he published the formula that must be properly called Carnoy (which others call Carnoy 2), which includes alcohol:chloroform:acetic acid (in the proportions 6:3:1) and is a most used and recommended fixative for important tasks.

Farmer and Shove (1905) used for studies of mitosis, two formulas of Alcohol-acetic, which applying the nomenclature I applied before, should be called alcohol: acetic 16:1 and alcohol: acetic 2:1 according to the formulas provided by Gray, 1954. (I have had no access to the original article, and I depend on Gray (1954) not registering a 3:1 formulation)

Who knows why, and who was the first to be wrong, but some botanists started to name the 3:1 as Farmer (forgetting Shove, and the real formulae implied) while some others call it Carnoy 1... and to use, with those wrong names, the Clarke’s formula... and so they do enthusiastically to this day in books and scientific papers. A short review on your browsers, and reading a few pages, will confirm this. My browser denounces over 300 articles that cite it when I search for (don’t forget quotes) “Clarke’s fixative”, 470 citations of “Farmer’s fixative”, 4800 when I search for “3-1 fixative” but 198,000 when searching for “Carnoy’s 3-1”. (Tell me, who said “Crime does not pay”?)

And at last, what is the Clarke? 

Clarke’s Fixative - (aka Farmer, aka Carnoy 1, aka 3:1) was introduced 160 years ago, in 1851, as a fixative for parts of the nervous system. Old as it is, simple as it is, it is good enough to be currently used professionally in the fixing of animal and plant histological pieces, especially for the study of cell division by the mechanisms of mitosis and meiosis: it fixes very well the chromatin.

It is also interesting that it was the first published formula for a histological Fixative. Its formula is

95 or 96% Alcohol (not less)   3 parts            75%

Acetic acid, pure                    1 part             25%

The "not less than" indicates that really the original formula of Clarke used absolute alcohol. But, if it is now difficult to find 96% alcohol, 100% alcohol is currently unreachable, for its price, and is indeed relatively useless for an amateur. Just open the bottle and the ambient humidity starts to contaminate absolute alcohol, which is highly hygroscopic. I could indicate ways to keep dry the absolute alcohol, or even how to prepare it from 96% alcohol, but they are not very easy, and for normal amateur applications, the 4% of water that has 96% alcohol does not bother at all. For this same reason, many professionals, reasonably, use this formula prepared with alcohol 96. (you know... 96% alcohol  is (almost) 96 ml of absolute alcohol + (almost) 4 ml water for every 100ml). This is a well known azeotrope mixture, and when alcohol is concentrated through distillation, alcohol cannot be concentrated purer than this.

 Even though this mix could only lead to Bolles-Lee’s contempt, the combination of the two items produces a fixative that, especially geneticists, have found extremely useful.

 See why...

 High purity alcohol precipitates proteins and is a very fast dehydrator. Its action is equivalent to suddenly drying the cells. The result is that, by inducing water loss, alcohol shrinks cells and tissues and harden the fixed pieces. This is especially important when you fix animal organs, because the tissue relationships can be easily distorted, and artifacts, especially non existing gaps and holes can appear. With the rigid walls of the vegetable cells, this could produce some sort of plasmolysis (but I don’t register it in my trials with onion epidermis, see the previous article).

 Acetic acid, on the contrary, swells, expands, cells and tissues, counterbalancing alcohol action, and precipitates DNA efficiently, making it an excellent chromatin fixative, encouraging the colouration of the nucleus. This is why both zoologists and botanists use it in the traditional histological fixatives.

 Thus Clarke’s appears to be, according to a more than a century-old experience, a quite balanced formula, in which each component controls the other component effects; that quickly permeates tissues (testicles, or anthers, for example, or even soft insect bodies); very easy to prepare, and long lasting if it is enclosed in a very good capped bottle. But... is so easy to mix in a hypodermic syringe 15 ml of ethylic and 5ml of acetic, that is better for me to prepare it in the moment.

 Probably therefore, I finally arrived at an acceptable scientific approach, and will use Clarke’s (1851) fixative, to see if it responds to the expectations and does a good job (a better job) with our epidermis.

 But see, and take note of the proportions, because the acetic concentration is very far from the dilutions Flemmings recommends, and that I tested still now!!! Vinegar is not an option in this formula, because of its high water proportion and pure acetic acid is not a cheap reagent.

 I immersed the epidermis for 4 or 5 minutes in several millilitres of fixative, in a capsule with lid (alcohol is a volatile liquid); the skin is so thin that this should be time enough (For histological frozen sections, which of course are thinner than the skin, approx. 1 minute is recommended). 

 But don’t feel hurried, at least animal pieces of tissues, or flower anthers, can be left in the Clarke’s for 24 or more hours, without damages.


Post-treatment of pieces fixed in an alcoholic fixative


As the fixative uses alcohol with a concentration of 96%, when I withdraw the epidermis off the fixative I wash it in 96% alcohol (a couple of minutes) and then in alcohol 70% (another two minutes). What I try is not to distort the fixed cells with strong and sudden changes of alcohol concentration... and, also, not to carry over acetic acid.


Fig 1 - This is my series: Clarke’s (with lid), 96% alcohol (with lid), 70% alcohol, collector flask (with tight lid)

 Then, this is a short series ... And let me introduce a novelty:  I can collect in a new volume of 70% alcohol, all the “peelings” I want, always maintaining a high proportion of alcohol to skin volumes. And besides, if I close and seal the container properly to avoid alcohol evaporation, I can keep the material for many weeks, perhaps months, having it available when I need it.

  It is an important advantage of working with 70% alcohol. Because of its water proportion it doesn’t make brittle the epidermis as 96% can do in the long term, and doesn’t allow bacterial development, which 96% can do.. It seems a paradox, but those differences of action are related to a different mode of interaction between the two alcohol’s degrees and the membrane of bacterial cells. 70% alcohol is a better bactericide, and so a better preservative.



It is also a norm that dyes should be in the same alcohol graduation than the piece to be stained.

 If I had the peels in 70% alcohol, and my “blue 1” solution is in water, I need to hydrate the epidermis.

 So I use the following series:


Fig 2

 Of course, if I start from scratch, the series I use is this

 Clarke >96% alcohol>70% alcohol>35% alcohol>water> “Blue 1”> Washing>rinsing

 I pass the epidermis through the selected series, taking them by their handles with my tweezers, giving two or three minutes in the intermediaries and 2 minutes in the dye.

I mount in water, as I do all over this series. See these example pictures.

epidermis 4x

Fig. 3 – Clarke – 4x obj

clarke's 10x

Fig 4 – Clarke – Blue 1, 10x obj.

clarke's 40x

Fig. 5 – Clarke – Blue 1 – 40x obj

Most notable at this magnification, is the inclusion in the mostly homogeneous cytosol of numerous dark granules, which are identified at a higher magnification as small dark spheres, different and distinct from the finer grains you see in the cytoplasm bands. Several preparations, from different onions, show the same.

By its disposition and size it's a big temptation to think they are "Golgi bodies". But without other techniques at hand to certify this I only can dream about it. They are some fluorescent images that induce my dream, but...


nucleus 6anucleus 6b


nucleus 6cnucleus 6d

Fig 6 – Clarke – "Blue 1" - Nuclei, 100x obj. Wetmount, Staining 5 min. See the “spheres” in a, b and d, and the defined bands of cytosol sown with dark granules in c.


 Chromatin is finely granular, nucleoli look good. In a couple of images, the colour of the nucleoli, verified by direct observation, tends to a dark reddish tint. In the cytoplasm, cytoplasm bands are confirmed. Nuclei, however, appear as discs in which they are noticed with difficulty the grooves which we know well from other fixatives, and which are a normal characteristic in the onion, as confirmed by this work

 In addition, direct observation shows that nuclear staining is not as intense, or of the same colour as on previous occasions. I believe that the difference is due to the high concentration of acetic acid. Manufacturers recommend the use of the Blue 1, mixed with vinegar, for the staining of "Christmas eggs" shells. So, one of my first attempts was to use the dye as they recommend. In my epidermis I got the same feeble tint result, and decided to use blue in aqueous solution without any addition. Clarke’s is a special case. Has a high acetic proportion. The images shown here are, of course, from cells where the nucleus were well coloured.

 Just as with high degree alcohol, the cytoplasm appears quite structured. (See fig 6, specially 6c)


Background coloration

 It seems that it's time to resort to an old invention of the histologists: background coloration with some dye colour intended to stain the cytoplasm with a different hue from the nuclear dye. 

 The oldest one, which has become a classic, is eosin.

 Eosin, discovered in 1871, was introduced in the histology techniques by Flemming in 1875. You see, this was a most fertile era for the experimentation on histological techniques.  One year before, Caro had discovered the Methylene blue. From the beginning (e.g. E.A. Schäffer, Practical Hystology, 1877) eosin was considered to be a generalized dye and was intended as a background for the contrasting nuclear colour, more dark and more chromatin specific, (eg. Moore (1882): for amphibians blood he uses methyl green for the nucleus, and eosin, for cytoplasm)

 Schäffer, even in the ’77, was already speaking of Lockwood-eosin, i.e. the predecessor of the well known Hematoxylin-Eosin.

 To clarify this item, see this Y. Lindekens work   l

It seems that eosin is more or less easily at the reach of amateurs. I buy it in a drugstore at Durango, and it seems to be sold in Europe as a Pharmaceutical 2% solution as a disinfectant.

 I use a 0.5% aqueous solution, applied after rinsing the “blue 1”, and for only 5 or 6 seconds. It’s a fast and tenacious dye.

 And then pass the skin to water, and to the slide, cover, and go to the microscope.

 The first unexpected result is that the eosin completely shifted the blue coloration of the nucleus. The final coloration is monochromatic and nuclei appear well coloured. (fig. 5 - 100xOI obj)(See my April 2006 Micscape article.)

 eosin 100x

Fig. 7

 The cytoplasm of the onion epidermis cells does not always have the same structure. In the epidermis of some onions, it is possible to distinguish the trabecular structure shown in Figure 8

eosin 40x

Fig. 8 - Fixation with Clarke and staining with eosin shows, in some onion epidermis, a detailed cytoplasmic structure. In picture above (taken through the 40x objective), we are lucky that there are two cells focused on different levels of depth. The lower, to the left of picture, shows the cytosol layer near the cell wall, forming a fine mesh of delicate trabeculae. The other, located immediately above, displays a classic median optical cut, with bands or trabeculae that cross the vacuole and bind the nucleus to the parietal cytosol. An interesting detail that we saw earlier in the test fixation with 96% alcohol, is the accumulation of cytosol and granules in cell angles.

 In others, the cytoplasm appears poorly granulated, and justifies or repeats many of the images that we have shown in cells stained with blue 1 in the preceding articles.

40x, eosin

Fig. 9

However who will undertake the work of reviewing them, will often see that many times before, I also presented evidence showing a reticulated or trabecular structure. Fixation with iodine (see 1st. part) and fixing with alcohol 96 (see. 4th part), as well as fixing with citric acid (see fig 6 and 9 of the 3rd. part) showed glimpses of what the eosin appears to confirm now in some skins, fixed with Clarke.(fig 8)

But most of the images with fixatives tested before show a rather homogeneous cytoplasm, and fine or medium granules, mostly isolated, or gathered in spots of different sizes.

 Here the key word is “some”, which easily translate to “not all”

 And then... which shows the true structure of the cytoplasm in the onion cell?

Many scientists, over many years had complained about the difficulty to know if what they see in their fixed and stained preparations had something in common with the live organism they started with. Bolles Lee (“the Bible”) put it in this way:

 From the practical standpoint, checking fixation images in living material is often difficult, and the method is limited in its application, and it may be asked by what criterion we may judge the quality of fixation. Unfortunately, there is no objective standard recognised and the matter becomes subjective. The careful investigator will try out a number of fixatives on new material, in order to eliminate any features which result from the use of a single reagent. 10th ed. Pg 23

 We (well... the professional histologists, of course) have now some very powerful tools to interrogate live cell’s nature, and can obtain some interesting answers. See this for example:

 Fig 10 – Arabidopsis sp. cells, Fluorescence images, modified from the on line article linked below


"Copyright American Society of Plant Biologists.”

The images portray the many paths actively built by actin (a linear, filamentous protein) bound to the cellular walls, coiled around the nucleus, and crossing through the central vacuole of an "arabidopsis" cell. Physiological and chemical studies had revealed that cytoplasm granules (mitochondria, Golgi bodies, “vesicles”, and other granular materials), travel along these pathways driven by “motors” of myosin.Oops! this is a deep insight, and a modern model of the live dynamics of vegetal cells!

 The same phenomenon is depicted in many images of “tobacco cells” found on the Net.

This is a good explanation of the phenomenon known as “Cytoplasmic Streaming” checked by kids, in school labs, from many decades ago, for example, in elementary biology courses, watching the movement of the chloroplasts in the cells of the leaves of "Elodea" or "Vallisneria".

 So, we can think that it is a generalized behaviour of plant cells, as you can also remember very old drawings of “Tradescantia” cells I share in the preliminary articles to this series.

 So... Do now I really know what I wish to find in my cells, fix with my reagents, and show in my slides?

 I think, my answer is yes, but...

 Have I shown this structures in the onion cells which I fixed and stained up until now?

 Well, for this first question I feel more or less comfortable... neither did my venerable old predecessors for more than a century. But... I must recognize that they laid the groundwork to make possible that today someone could show this phenomenon. The above fluorescence picture speaks for itself. And, reviewing my pictures I think  that hints of this structures show in some of them.

 Is this structure permanent and always detectable? Is the living onion cell alike with the ones with a reticulate cytosol that fluorescence images of cells from other plant species show, or are more alike to the more numerous examples I grab from unstructured onion cytosol pictures?

 For this second question I have seen very scarce and partial answers on books or the Internet. I think I must discover this with my own onions. It is very interesting. The answer to what I must search for (if I have the tools to do the task) depends on this.

 It is clear that I must forgive the fixatives, primitive and simple, I used in my first attempts, which did not show reliable images of such a complex structure. But... Should not a more respectable fixative as Clarke’s give us a more “close to life” image, i.e. an image most closer to the fluorescence image?

 Perhaps I need to leave the task to those who have the advanced and costly optics and reagents that could achieve the job. Or use other fixatives, or dyes. And have a better equipped laboratory. Or all of these!

Perhaps... Perhaps... This sounds disappointing for me.

 But... In Spanish we have a “proverb” that says “ no está muerto quien pelea” This translates in English into something like: “It’s not a dead man who still fights”

 I wish to give some more fight... next month, when I recover from my desperation.


 INTERESTING: I had forgotten. NO, NO BUBBLES, not even one, when using the Clarke.


 The only additive to the acetic acid (which itself was ineffective, used alone, as for the July’s article) is the high degree alcohol, which also could not do the job alone, as you see in the August article. I think I must give the merit to “the formula”!!!! Thanks, J.L. Clarke!



Comments to the author, Walter  Dioni , are welcomed.

Micscape Magazine
Article Library
Microscopy UK Front Page


© Microscopy UK or their contributors.

Published in the September 2011 edition of Micscape.

Please report any Web problems or offer general comments to the Micscape Editor.

Micscape is the on-line monthly magazine of the Microscopy UK web
site at

© Ltd, Microscopy-UK, and all contributors 1995 onwards. All rights reserved.