To prove the existence of biological transmutation, experiments must be bound by spatial and temporal considerations.

Living organisms are in constant exchange with their surroundings. To record the changes in chemical composition that occur in a particular amount of material which is under the influence of a living organism, one is committed to a quantitative determination of all the elements inside a closed space (hereafter known as the system), that encloses both the living material and a piece of its natural environment. This environment includes, apart from light as an unweighable component [though the mass of light is practically unweighable, it's energy is measurable, and since it is critical for the development of all green plants, full details of the intensity and nature of any light source must always be recorded (see for example Tipnis and Pratt, 1960) - a point neglected in his German language article], the air gasses, water and salts [; the system should also include any solid surfaces, such as the bounding walls of the experimental system which, though chemically non-reactive, may ab- or ad-sorb a proportion of one or more of the above]. A continuous provision of the gaseous component is necessary for the maintenance of life, thus the quantities of the gasses consumed and formed must be determined and ideally one should work in an air-tight system. Although one such attempt has been made (see section 4.3.) an experiment conducted in such a way would not be easy, so that, despite working towards perfection, one would generally limit oneself to non-volatile elements and those which do not volatilise during ashing. [Ashing is a process here involving the heating of a sample to about 500 degrees in an oven, thus burning off all organic matter leaving only an inorganic, mineral residue]. Such non-volatiles can be studied, in principle, in open containers without losses. For the further enclosure of the system, clear (glass or plastic) lids can be used with openings for the entry and release of gasses.

By such a choice of system, one loses very many possibilities that an experiment may have to offer. Should it be possible for a transmutation to occur with one of the four "organic" [my emphasis] elements (carbon, hydrogen, oxygen and nitrogen) [all of which may be volatile as element or compound], this could not be determined by such a limited experimental design.

[Holleman notes in the margin here an experiment by Picket on the measurement of argon gas production by yeast.]

For an absolute proof of a change in the composition of a system an analysis must be conducted of this system in at least two stages (a beginning and an end stage), over a period in time during which a development in the organic material may occur. This is in practice only to be used as a method when an arbitrarily formed change in the composition of an, in part, living system can be followed without profound interference in the structure of its matter. Existing methods for this ([left blank but NMR spectroscopy might be one such method]) have not yet been worked out in a form that lends itself to the research that is being considered here. The application of the standard methods of plant analysis, which result in the destruction of the plant material, are totally unsuitable to the research described above.

One is therefore forced to conduct this research on the beginning and end stages of two or more independent systems [that are in all relevant aspects identical]. From a part of the plant material that is as homogeneous as possible, in the initial stages (seeds, roots, etc.) one can put a number of systems of the same type together; some of which (the control systems) are directly analyzed [see however section 8.1.5.3], the remaining (the experimental system) only after the development of the plant.

This method has been, until now, generally used in biological transmutation proofs, often with the neglect of the statistical problem that arises here. The composition of different plant structures, e.g. leaf or stalk, also that of a particular quantity of a determined weight of seeds from a single lot, is in general fairly variable [Holleman speaks here from personal experience based on his early unsuccessful work to repeat Herzeele's extensive transmutation researches]; with experiments involving cuttings, variations of 10% or more are no exception. Such variations even occur when each half of a longitudinally divided stem is tested. The question as to whether an experimentally determined difference in composition between experiment and control is significant can nowadays be answered by well known statistical methods. Because no such account was given with older research, the results obtained were often illusory [Holleman refers the reader here to an intended appendix which was never written, describing his earlier Herzeele research].

Significant results [positive or negative] can in these cases always be obtained by increasing the number of replicates. Should the starting material exist in smaller units (e.g. small seeds) it is to be expected that it will be more homogeneous, so that significant results may be obtained with fewer replicate experiments.

The statistical problem is totally lost with the use of single celled organisms, e.g. algal suspensions of a microplankton type (average cross-section of the cell a few hundredths of a millimetre) [provided that the cell population is sufficiently homogeneous]. When sinking is avoided, variations in the average composition of such suspensions apparently fall within the errors of the experimental method.

The question as to which point in time such an experiment can be shown to have ended is only to be answered in relation to the degree of development that the organisms attain during the experiment. An exploratory piece of research on biological transmutation should in the first instance be directed towards the tracing of an effect within the whole course of development of the organisms. Now, in the case of working with larger plants, the development of the different individuals in a system is not uniform; only some (in the most favourable, the majority) go through the whole development expected under the given conditions. Thus chemical analysis of the system as a whole always delivers an average of all the attained development stages. This applies probably just as much for a culture of microorganisms.

So long as the development of the system is clearly in an on going direction and the finishing material is of reasonable homogeneity, it may be pretty well assumed, that a large majority of organisms are associated in a comparable stage. Under these preconditions the best chance of reproducing the outcomes also exists.

The best moment for the ending of the experiment may be provisionally chosen when it is best estimated that the maximum unfolding of the organic life is attained, with the proviso that also in the break-down and decomposition stage, effects can still occur.

So long as one is unable by any means to determine in which development stage it is possible for an organism to produce a transmutation effect, it is necessary not only to arrange an analysis at the begin and end points of the system, but also at different points in time during the experiment. Accepting that at each point in time a majority of the individuals of a system occur in approximately the same stage of development, one can hereby in any case get an impression of the progress of the process. Things that occur, for example, in a particular stage of the process that become compensated for by their opposite in a later stage, can only, by this means, come to light.

For the practical conducting of this research one is driven to the setting up of a number of parallel experiments with systems of identical composition, that each may be broken off at a different point of development.

Since, for the research described here, the specific influence of the organism's life process is at issue, the extent of the organic development must, within the accepted time limit, be promoted as much as possible.

With an eye to this the following measures are put forward.

The supply of the essential mineral food components in optimal quantities is probably an important factor for the correct unfolding of the organic process. It is noteworthy that earlier researchers paid relatively little attention to this. The material used was allowed to germinate either in distilled water or supplied with extremely one sided nutrient solutions. The reasons, which were given earlier, shall only quickly be gone into in the following ( )[sic]. In principle it is not to be discounted that only in deficient nutrient solutions does a plant use a transmutation process as a way out.

For an exploratory piece of research the choice should really be given to the optimal feeding of the experimental organism by means of the organism's most suitable nutrient solution. Hereby the material transmutations during the progress of the organic process can be studied and all phases of the development will be obtained. [See however section 10.1.3].

[This section was compiled from two rough pencil written drafts of Holleman's; I am therefore responsible for a minor amount of editing to provide a single, hopefully coherent text. The principle here described, though essentially a simple one, has not proved easy to explain; it is however of crucial significance to Holleman's experimental design. Nevertheless I have refrained from rewriting Holleman's original text. The reader is thus referred not only to the last paragraph of this section, but also to section 6.1 and section 7.1.1 for practical examples of how the cumulative method was conducted.]

The experiment shall generally be set up so as to attain the largest possible effect. An as yet untested means towards this end is the single or multiple replication of the growth experiment in one and the same system [see section 3.1.1 for a reminder of the definition of a system as used here]. This means of working appears specially suited to experiments with a small amount of homogeneous culture material in which a noticeable development can be expected, i.e. for example, microorganisms. A possible transmutation becomes, in the first instance, just as many times increased as the number of part experiments. The accumulation in effects possibly limits a natural growth when, by this method of working, an essential element is removed from [or added to] the nutrient solution in too great a measure.

At the end of the experimental period in the first experiment, one ashes the content of the dish system and brings these back again [by dissolving the ashes] into the original form of the nutrient solution that was chosen [; I have expressed doubts, in section 10.1.5.2, as to whether the exact original chemical composition of the nutrient solution was in fact reconstituted]. This should generally be possible without adding, or losing, measurable amounts of ash components (cations) [I do not understand why anions are not also considered important; of note here is that Holleman found it necessary to add extra nitric acid to adjust the pH of the reconstituted nutrient solution back to it's original value]. The reconstituted nutrient solution obtained, can be used again for a growth experiment in the same system. [These guidelines should obviously be considered as just that; the results of any modifications should always be measurable and should not affect the health of the cultured organisms.]

(The following text only in key words; to be elaborated later.) [Sic].

The avoidance of positive contamination [causing artificially high results] and negative [causing artificially low results] contamination. Culture containers must be completely inert over a long period; resistant to long term action of neutral watery solutions, also against corrosive chemicals (nitric acid, perchloric acid) and against heating to ca.500 degrees (ashing). Quartz glass is the best. Protection against infalling dust. For the growth of algae, agitation of the solution is necessary: shaking machine at constant temperature [water-bath]. Supply of air carrying 5% carbon dioxide through flexible tubing and, by means of a splitting point, equally divided over the culture dishes.

Parallel working of control and experiment; the handling must be identical, with the exception of the addition of an excess of nitric acid, with the experimental cultures at the end of the growth period, with the controls directly at the setting up of the experiment, so that the organic development is prevented [this was the unsuccessful practice used only in experiment III; see section 6.1].

Working in as sterile conditions as possible.