Based on the extremum seeking (ES) algorithm, an adaptive scheme for nonholonomic source localization in a 3-D environment without position information is developed. While the previous works use the sinusoidally perturbed form of ES, this technical note employs the recently form of ES in which the measured quantity for optimization appears as the argument of a sine/cosine function. The proposed scheme takes advantage of the fact that the key part of the recently form of ES algorithm (the sine/cosine function) already exists in the vehicle model. Stability analysis of the proposed scheme is performed using the averaging theory and its performance is illustrated with simulation examples.