Let us consider, for example, the case of a product which gives out only α rays. The passage of the α particles through the gas produces a large number of ions in its path. Since the α particles from any individual product are projected with the same average velocity under all conditions, the relative amount of the ionization produced per second in the testing vessel serves as an accurate means of determining the variation of its activity. No two products, however, emit α particles with the same average velocity. We have seen that the rays from some products are more readily stopped in the gas than others. Thus the relative saturation current, due to two different products in a testing vessel, does not serve as an accurate method of comparing the relative number of α particles expelled per second. The ratio of the currents will in general depend upon the distance between the plates of the testing vessel, and, unless the relative ionization due to the average α particle from the two products is known from other data, the comparison of the currents can, at best, be only an approximate guide to the relative number of α particles escaping into the gas. =202.= Some examples will now be considered to show how the factors, above considered, influence the character of the curves of activity obtained under different experimental conditions. For the purpose of illustration, we shall consider the variation after removal of the excited activity on a body exposed for different times to a constant supply of the radium emanation. The active deposit on removal consists in general of a mixture of the products radium _A_, _B_, and _C_. The nature of the rays from each product, the time for each product to be transformed, and the value of λ are tabulated below for convenience:— Product Rays T. λ (sec⁻¹) Radium _A_ α rays 3 min. 3·85 × 10⁻³ Radium _B_ no rays 21 min. 5·38 × 10⁻⁴ Radium _C_ α, β, γ 28 min. 4·13 × rays 10⁻⁴ Since only the product _C_ gives rise to β and γ rays, the activity measured by either of these types of rays will be proportional to the amount of _C_ present at any time, _i.e._ to the value of _R_ at any time. For a long exposure, the variation of activity with time measured by the β and γ rays will thus be represented by the upper curve _CC_ of Fig. 73, where the ordinates represent activity. This curve will be seen to be very similar in shape to the experimental curve for a long exposure which is given in Fig. 68.