Some aspects of a technique for carrying out and interpretation of smoke experiment results in the Big aerosol chamber are discussed. The experiments have shown that despite sufficiently stable conditions for development of smokes in the chamber, the smoke microstructure undergoes essential qualitative and quantitative variations. A closed numerical experiment has shown that the adequate interpretation of the spectral nephelometric data strongly depends on the accuracy in preestimated mean-effective complex index of refraction of smoke particles. Analysis of some results allows an assumption that the used method gives an adequate estimate of the efficient index of refraction and particle size spectrum in the presence of a random instrumental error not exceeding 10-14%. Estimates of efficient values of the index of refraction show that there can be both weekly absorbing particles and particles close to soot by their characteristics. In the dynamics of the smoke particle characteristic sizes, there appear different trends at different intervals of size scale, which are due to appearance of different proportions in the content of particles of the intermediately dispersed fraction.