Introduction. As demonstrated elsewhere ¹, the developing eggs and embryos of a fish, Misgurnus fossilis L. are interacting with each other in such a manner that the embryos of more advanced stages are in most cases suppressing the development of younger ones. This peculiar event should be linked somehow with the biophoton emission of embryonic samples. However, the classical approaches are not working here because the average amount of the biophoton emission from embryos do not exceed significantly the background level ² and is therefore from the classical view "nothing". But how "nothing" can induce anything?

We tried to explore this problem by using more refined methods of a statistical analysis. Firstly, we employed what we call a "zero-one" statistics, that is, we count a number of zero, 1 and in some cases few more counts per small dwell time period (which we took in these experiments as 0.1 s) (Fig. 1). Then, we have used Fourier analysis with some modifications. Namely, we have measured the Fourier spectra (FS) in terms of spectral densities for brief enough successive time periods, each of them consisting of 1000, 2000 or 8000 successive measurements (which corresponds, under 0,1 s dwell time, to 100s, 200 s or 15 min time duration). Then, for getting more condensed presentation of FSs, we calculated the autocorrelations within the spectrae under a substantial number of lags (in most cases we took 300 lags covering the range of 0,2 – 60 s oscillations periods). We define the obtained autocorrelations diagrams as AC-FS patterns. And, finally, we took the successive partly overlapped periods of UWPE records and looked for the correlation patterns of the records themselves and/or their FS. In most of these studies we took the periods of 1000 measurements each (=100 s) overlapped to 10%.

Developmental dynamics of the biophoton emission in terms of "zero-one" statistics. We could detect a significant and regular increase of zero counts during the initial period of development, starting from non-fertilized eggs up to the end of cleavage divisions (Fig. 2). At the end of this period the amount of zero counts reaches the background level and keeps this value until the end of development (including larval period) with only one possible exception (an object for further testing!) at stage 31 which corresponds to a hatching period (Fig. 3). In any case, the embryos from stage 11 up, which are the most active partners of the optical interactions emit in the average virtually nothing! Do they show, meanwhile some specificity in the terms of oscillations of a biophoton emission intensity?

Developmental dynamics of the biophoton emission in terms of Fourier patterns. As seen from Fig. 4 (left column) the Fourier spectra of the biophoton emission from the different stage embryos are indeed different. This can be made more representative if using, instead of the spectra themselves, the autocorrelation diagrams (AC-FS patterns). As one can see from Figs. 4 and 5, these ones can be divided into the following categories:

1. Chaotic patterns: most of the AC coefficients are below the significance border;
2. Unstable patterns: AC coefficients are extensively changed even after 2 min time periods;
3. Stable periodic (comb-like) patterns which may have different values of the AC maxims and different distances in between; correspondingly, one may distinguish a more fine and a more coarse periodicity;
4. Gradual (slope or zigzag-like) patterns, characterized by much more extended spectral densities maxims. Hence, such patterns possess a long-range order lacking in the previous group.

As shown by Fig. 5, the upper row patterns (records of photon emission from a quartz cuvette with water but without embryos) are practically non-significant. Non-fertilized eggs exhibit some quite unstable patterns, sometimes with highly extensive AC values (frames 2 and 5). The third row from above illustrates quite remarkable (in terms of UWPE patterns) brief developmental between fertilization and start of the cleavage divisions. Here one can see very pronounced and rather stable comb-like periodic patterns without any signs of a long-range order. Somewhat later in development (blastula stage, fourth row) the AC coefficients are generally decreased, but the comb-like patterns still preserved. Even later the AC coefficients are even more decreased but, instead, the remarkable long-range order patterns (zigzag-like or monotonously sloping) are emerged. That means that the local spectral peaks are to a great extent exchanged by more smoothed areas of increased spectral densities. In any case, there are definite age-specific differences in Fourier patterns which may be related to the stage-dependent effects of the optical interactions between embryos. This conclusion is confirmed by a comparison of the cross-correlation diagrams of the Fourier spectra for the different stages of development. As one can see from Fig. 6, the diagrams for the cuvette+water controls, non-fertilized eggs and those reaching stage 11 are rather different. While the cross-correlation curves for the successive states of the cuvette+water controls are mixed in a random manner, those related even to non-fertilized eggs (not to say about developing embryos) show a greater order being subdivided to prolonged periods of a relatively high correlation and rapid transient periods. This is espacially clear for developing embryos. We got many examples of such a kind, Fig. 6 being just one of them.

Optical interactions between embryos, statistically evaluated. While measured in terms of "0,1" statistics, the optically interacting batches (placed in small quartz cuvettes, staying close to each other before PMS window) show a gradual increase of zero counts (Fig. 7, interact). Same batches screened from each other by a piece of a black paper do not show this effect (Fig. 7, sum). These measurements have been performed on 12 pairs of interacting batches. The results are statistically significant (Fig. 8). Similar results, that is, a gradual decrease of a total biophoton emission of the optically interacting samples as compared with the same samples which do not see each other could be also observed under greater dwell times (Fig. 9. 10).

On the other hand, during optical interactions we have observed an increase in kurtosis (this is 4^th order momentum, a sensitive indicator of the amount of high peaks: see STATISTICA program, Quick Basic Stat). When the batches of the different stages have been brough into optical contacts, the kurtosis increased mostly in the more advanced batches and the greater was the age difference of the interacting batches the greater was this increase (Fig. 11).

Brief discussion and conclusions. Firstly, we would like to emphasise that the amount of a spontaneous biophoton emission (in terms of the increase of zero counts) decreases as a development proceeds. This seems to be a general rule as being shown beforehand on the muscles of newborn rabbits ³ and in chicken embryos ⁴. What is new in the results above presented is that in the case of the fish eggs such a decrease is closely correlated with the increase in DNA/cytoplasmic ratio. This may be considered as (up to now) indirect confirmation of the hypothesis of the role of DNA in photonic storage 5,6.

Other property which may be of a biological importance is the change of Fourier patterns of a biophoton emission within the course of development. These changes are going on in a holistic manner, that is, involving wide spectral parts rather than some individual spectral maxims (see, e.g. Fig. 4) and being manifested in the increase of the high correlation periods (Fig. 6). In particular, a regular extension and smoothing of the spectral maxims in the advanced embryos indicates, in the formal terms, an increase of a "friction" between the "neighboring" oscillators, as a development proceeds. A similar phenomenon of enlarging and fusing together some spectral lines was described ³. Further investigations are required for specifying this important matter.

A decrease of a biophoton emission (Figs 7-10) during optical interactions of embryonic batches can, by our view, be interpreted only in terms of a destructive interference of the coherent electromagnetic fields ⁷. On the other hand, the increase in kurtosis may point to the constructive interference. In any case, both the developmental dynamics of a biophoton emission and its changes during optical interactions cannot be understood in the classical terms of the emission and exchange of a photonic energy (because just during the mostly active optical interactions the total amount of this energy dos not exceed the background level!). The only possible way for explaining them is, by our view, to assume that the embryos of a single batch and of the optically interacting batches create a common coherent electromagnetic field providing a "photon sucking" ⁷.

The authors express their thanks to Professor F.-A. Popp and to the collaborators of the IIB for their hospitality, valuable discussions and technical assistance.

References

1. Burlakov A.B. (2000) Distant physical interactions between the developing fish embryos // Biophotonics and Coherent Systems (L.V.Beloussov, F.-A.Popp, V.L.Voeikov and R. Van Wijk eds). Moscow University Press. Moscow. P. 289- 304.

2. Beloussov L.V., Burlakov A.B., Konradov A.A. (2000) Biophoton emission from eggs and embryos of a fish, Misgurnus fossilis l.: developmental dynamics, frequency patterns and non-additive interactions. // Biophotonics and Coherent Systems (L.V.Beloussov, F.-A.Popp, V.L.Voeikov and R. Van Wijk eds). Moscow University Press. Moscow. P. 305-320.

3. Gurvich, A.A. (1992) Mitogenetic radiation as an evidence of nonequilibrium properties of living matter. In: F.-A. Popp, K.H. Li and Q. Gu (eds) Recent Advances in Biophoton Research and its Applications. World Scientific. Singapore etc. 1992 pp. 457-488.

4. Beloussov, L.V. and Louchinskaia, N.N. (1998) Biophoton emission from developing eggs and embryos: non-linearity, holistic properties and indications of energy transfer. In: Biophotons (Jiin-Ju Chang, J.Fisch and F.-A. Popp eds) Kluwer Acad Publ. Dordrecht/Boston/London P. 121-142.

5. Li, Ke-hsueh (1992) Coherence in physics and biology. In: Recent Advances in Biophoton Research and its Applications (F.-A. Popp, K.H. Li and Q. Gu eds) World Scientific. Singapore etc. P. 113-156.

6. Niggli, H.J. (1992) Biophoton re-emission studies in carcinogenic mouse melanoma cells. In: Recent Advances in Biophoton Research and its Applications (F.-A. Popp, K.H. Li and Q. Gu eds) World Scientific. Singapore etc. P. 231-242.

7. Popp, F.-A. (2000) Some features of biophotons and their interpretation in terms of coherent states. Biophotonics and Coherent Systems (L.V.Beloussov, F.-A.Popp, V.L.Voeikov and R. Van Wijk eds). Moscow University Press. Moscow. P. 117-134.

Fig.1. Graphs of "0,1" statistics for several M. fossilis eggs’ samples (nonfert: non-fertilized eggs; 60 min: 1 h after fertilization; 2 blast, 16 blast: 2 and 16 blastomeres, resp.; water: cuvette+water controls; dark count: PMS background). Horizontal axis: counts per 0.1 s; vertical axis: percents of counts. Note, that except non-fertilized eggs, the samples differ from each other and from the control measurements only in terms of zero counts.

Fig.2. Dynamics of biophoton emission (in terms of "0,1"-statistics) for the initial period of the development of M. fossilis eggs (from non-fertilized eggs up to 165 min after fertilization). Vertical axis: percents of 0, 1 and 2 counts per 0.1 s. Note a gradual increase of zero counts.

Fig. 3. Dynamics of biophoton emission (in terms of "0,1"-statistics) for the advanced stages of the development of M. fossilis samples (embryonic stages from 10 to 31 and the larvae of 2.5, 4-5 and 10 days after hatching). Vertical axis: percents of 0, 1 and 2 counts per 0.1 s. Most of the values are at the background level. A drop of zero counts (i.e., an increase in biophoton emission) at stage 31 is probably associated with hatching.

Fig. 4. Relations between the autocorrelation (AC) patterns of the Fourier spectra (left column) and the spectra themselves (right column) for several successive developmental stages of M. fossilis samples. Note the "comb-like" AC patterns typical for early embryos, as opposed to the wavy and zig-zag patterns emerging at the later stages.

Fig. 5. Sets of representative AC-FS patterns of a biophoton emission illustrating their developmental dynamics in M. fossilis samples (stages shown at left). Upper row: cuvette with water only. Each row frames correspond to 5 successive 2 min duration records of the same sample.

Fig. 6. The dynamics of the correlation coefficients (CC) for the raw data and for the Fourier spectra of a biophoton emission in the 10% overlapped records of non-biological controls (cuvette with water, left column), non-fertilized fish eggs (central column) and stage 11 M. fossilis embryos (right column). Upper row: records of a biophoton emission per 300 s (dwell is 0.1 s). Next raw to below: a set of CC for the raw data measured within successive 100 s periods overlapping each other to 10 s. Note the increase of CC values and the emergence of their periodicity in the right column. Nevertheless, even the highest CC do not exceed 0.1 (a verge of a significance). Raw third from above: similar diagramms for CC of Fourier spectra. Lower raw: mid-fragments from these diagramms. From left to right CC changes become more regular and cooperative.

Fig.7. Dynamics of zero counts under 0.1 s dwell times during the optical interactions of two batches of M. fossilis embryos ("interactions"), as compared with the same batches which do not see each other ("sum"). Horizontal axis: time of interactions (or staying together without optical contacts), minutes.

Fig. 8. Averages for 0, 1 and 2 counts per 0.1s dwell periods for 12 pairs of the optically interacting (int) and non-interacting (sum) embryos.

Fig. 9. Frequency curves for the biophoton emission of two optically interacting batches of M. fossilis embryos after 10, 20 and 30 min of optical interaction. Dwell time is 0.5 s. Horizontal axis: counts, vertical axis: percents. Note leftwards shift of the most frequent values with the course of interactions.

Fig. 10. Frequency curves for the biophoton emission of the same batches as in Fig. 10 which now do not see each other after 10, 20 and 30 min of staying together. No changes in the emission intensity.

Fig. 11. Changes in kurtosis as a result of the optical interactions of the batches with the different stage interval.