In many rock magnetic studies, information on magnetic mineralogy is of crucial importance. Besides standard analytical methods, such as X-ray spectroscopy, more sensitive thermomagnetic analyses are often used. Temperature dependence of magnetic parameters can serve as basis for determination of magnetic second-order phase transition temperatures. Although limited by several drawbacks, the most serious being thermally induced transformations of the original minerals, this method provides useful information not only about the presence of magnetic minerals, but also additional knowledge on, e.g., the prevailing grain size distribution or degree of substitution. In thermomagnetic analysis, temperature dependence of two parameters, induced magnetization and magnetic susceptibility, is mostly used. However, let us say because of historical reasons, the same approach for the Curie point determination has been often used in analyzing the two parameters. In our contribution, we discuss the physical principles of the two parameters, showing that the methods developed and used for induced magnetization cannot be used also for temperature dependence of magnetic susceptibility because there is no physical justification to do so. Otherwise, the error in determining the Curie point can be some few degrees but can reach also several tens of degrees. Such an error has serious consequences for further interpretation of the data, e.g., in terms of degree of Ti substitution in Ti magnetite.