Notes. # – number of data points; ⟨ μ ⟩  (β_app), ⟨ μ_lower ⟩ , ⟨ μ_upper ⟩  – average component speed in mas year^-1, in parentheses the apparent speed β_app, for Q9 and Q10 lower and upper limits of the component speeds; ⟨ Φ ⟩ , ⟨ Φ_lower ⟩ , ⟨ Φ_upper ⟩  – average position angle of the component motion in degree, for Q9 and Q10 lower and upper limits; τ_e (τ_lower, τ_upper) – Ejection epoch/time of zero separation from component labeled Q0, determined through linear extrapolation using the average component speed and component position at the midpoint, t_mid = 0.5·(t_max − t_min), of the respective dataset, for Q9 and Q10 lower and upper limits; S₁, S₂ – maximum, minimum (at the time of maximum absolute value of the time derivative) component flux density in units of Jy; d₁, d₂ – circular Gaussian-modelfit component FWHM in units of mas at the time of S₁, S₂; dt – time between S_max and S_min in units of years; Δt_var – time variability factor in units of years; δ_var, Γ_var, Θ_var – derived quantities of the variability Doppler and Lorentz factors, and the viewing angle to the line of sight in degrees (*) – indicator whether component was accelerating (yes) or not (no). The asymptotic standard errors in the fits were determined statistically using an implementation of the nonlinear least squares Marquardt-Levenberg algorithm with the position errors as weights. Errors in the derived parameters are 68% confidence limits determined through Monte Carlo simulations.