Observation, one of the methods of scientific cognition, is found to closely link with experimentation. It is accepted to relegate observation to the passive methods of cognition since observation does not produce interference in natural processes. This circumstance begets a distinctive relationship of observational material to theory. Theoretical concepts do not rise spontaneously from observational material. The grounds for the rise of a theory is practical action including scientific experimentation. Verification of the correspondence between theory and objective reality is brought about by means of a large complex of procedures, which mediates the relation of cognition and practice from technological experience to philosophical-gnosiological norms. The rôle of an observation, not attached directly to scientific experiment, is determined by this circumstance in that it is tied ambiguously to theoretical concepts. For the same observational material one may find alternative explanatory schemes. For example, philosophical and physical presentations of antiquity, as a rule, came from observed facts to which we now ascribe completely different explanations. In contemporary science one may also run into a situation where confirmation of theoretical postulates exists in an observation that may have alternate interpretations for its results. Such a condition took place, in part, in the observation of binary stars and their use to confirm the second postulate of relativity as a counterbalance to the ballistic hypothesis of Ritz.

In a discussion on the existence of the electromagnetic ether in 1908, Ritz advanced a hypothesis extending the principle of relativity to electromagnetic radiation. (Ritz's hypothesis was given the title "ballistic" because the emission of light was compared to the motion of a projectile fired by a moving gun.) In 1913 de Sitter introduced discussion about observations of the motion of visually-binary stars to refute Ritz's "ballistic" hypothesis, the essence of which follows:

". . . imagine a double star (Fig. 1) located at a distance L from an observer. One of the components S has an orbital period 2T and a linear speed v. If the "ballistic" hypothesis is correct, then the light from component S at position I will reach the observer at the moment t1 = L / (c - v) ; and from position II at moment t2 = T + L / (c + v) , where T is the semi-orbital period.

Thus the visible motion of a star may markedly deviate from Kepler's laws. In particular, when there is a very large L, it is possible that even with v much less than c one can get t2 less than t1, i.e. the visible motion acquires a highly whimsical character. The examination of a sufficient number of stars shows that such an approach as the "ballistic" hypothesis contradicts observation, and why, consequently, Ritz's hypothesis ought to be abandoned." [1]

But is this necessarily so?

Having offered the above reasoning, we may conclude that if there exist in the motion of visually-double stars assumed deviations from Kepler's laws as a result of the addition of speeds, they are so small that they cannot be registered by our instruments.

Figure 1

In order to show this, let us take the angle alpha between images of star S at points I and II under the condition that t1 = t2

(1)

Re-arranging and solving for L we get:

(2)

The distance from point I to point II equals the orbital diameter D across which the star moves in time T. This allows us to write:

(3)

One can see (Fig. 1) that the angle alpha, for D much less than L , equals tan alpha ; that is:

(4)

Substituting into (4), in lieu of D and L, their values from (3) and (2) and recalling that v is much less than c, we find:

(5)

It is known that the speed of visually-binary stars in orbit is much less than the speed of 350 km/sec, which is necessary for the angle alpha to amount to 2 x 10exp-6 radians, which is the resolution limit for present day telescopes. Therefore trigonometric measurements of the orbits of visually-binary stars do not allow us to have made any conclusion about the suggested hypotheses.

However, the result of adding the absolute speed of light and the speed of its source relative to the observer ought to reveal itself as changes in intensity of the received radiation, when compared to the constant intensity of the star S. This is so because at certain moments the "faster" light catches up with the "slower" light and is received simultaneously by the observer. This means that the speed modulation leads to an intensity modulation with simultaneous changes in the "observation" in step with the Doppler effects of the orbit. [Emphasis added. Trans.] In order to analyze the character of this occurrence, let us construct light trajectories with coordinates L and t (Fig. 2) having speeds C(i) = c + v sin omega t, traveling from star S which is moving in a circular orbit.

The intensity of radiation B of star S is constant. The intervals = 2 T / n are equal to one another [at L = 0 ] (n being an arbitrary number). At the distance L(1) let us label intervals , which are contiguous, and are between trajectories. We will find that the received stellar brightness for a given distance equals , (or if the light comes from several regions in the orbit). Let us label the distance as Lo and let it be our stipulated unit distance to the given star. This will be the distance for the simultaneous arrival of the light from points I and II in the orbits, for a given T and v.

The conditions shown in Fig. 3 are a portrayal of the dependence of the orbital speed and stellar brightness on time which ought to be observed at the distance 0.5 Lo, 1.0 Lo, and 1.5 Lo from the star. Similar characteristics are possessed by so-called "variable- pulsating stars" a hypothesis which resembles that of binary stars with only one star shining which was introduced at the beginning of our century.[2]

(1) The computed apparent radial velocity curves lag the brightness curves by 90 degrees in 3a, 3b, and 3c. (This is not obvious for 3b and 3c.) This phasing offset, at first glance, seems fatal to Sekerin's hypothesis.

(2) Recomputed brightness variation curves, furnished by Sekerin's colleague, M.S. Serbulenko, are shown for the second cycle. (Faster-later light overtaking slower-earlier light produces the unorthodox time reversals in arrival times in the brightness and apparent velocity curves in 3b and 3c.)

See A Ritzian Interpretation of Variable Stars for a resolution of the 90 degree phasing error. Copies of three of Serbulenko's light and apparent radial velocity curves are also reproduced and discussed.