Abstract
The accuracy of a wavefunction obtained by the author using a transcorrelated procedure is determined using a numerical method. The investigation shows that the wavefunction is of about the same accuracy as the corresponding variational wavefunction. It is shown that a slight modification to the author's calculation gives a wavefunction which is more accurate than the variational wavefunction. None of the transcorrelated wavefunctions examined here is as accurate, however, as the 12-term wavefunction of Kołos and Roothaan.