The writer has used Greek and Roman names for these gods to some extent indiscriminately, in accordance with the universal custom in art. Nevertheless the practice is to be regretted as it tends to complicate the general ideas of the Greek and Roman religions. Notwithstanding the occasional direct association of some of their deities with human personages by their poets, the Romans regarded their gods as purely spiritual beings, having no special earthly habitation, or sex relations with the human race, while their powers widely differed from those of the respective Greek deities with whom they are commonly identified. Authorities differ as to whether the gods were supposed to have spiritual marital relations with each other.[a] In any case the whole nature of their religion precluded the development amongst the Romans of a separate sacred art. Their sculptured gods, which were taken from Grecian models, were symbols rather than presumed types. [a] See J. G. Frazer's _Adonis, Attis, Osiris_, 1914, vol. ii.; and W. W. Fowler's _Religious Experience of the Roman People_, 1911. NOTE 44. PAGE 111 If we may judge from the headless figures of the goddesses, commonly known as the Three Fates, from the east pediment of the Parthenon, there seems to be little difference between the general lines of the feminine torso represented by the Phidian ideal, and those of the Praxitelean model. The Parthenon torsos are more massive proportionately, but the object of both Phidias and Praxiteles was evidently to straighten the outer lines of the torso as nearly as possible, making due allowance for the varied natural swellings of their respective forms. It is obvious that the use of attire gave Phidias (presuming the Parthenon figures referred to were designed by him, as they probably were) a latitude in varying the proportions of the torsos which he could not have exercised in the case of nude forms. Unclothed, the figures would appear unwieldy, and the graceful flowing lines resulting from the partly clinging drapery could not be so completely presented with nude reclining or semi-reclining figures. There are other features also which prevent the nude representation of such massive forms. Thus, the breasts would necessarily be out of proportion in size, and widely separated. These conditions are common in fifth century and archaic figures, and do not appear to be defects in forms of life size or less, but they would be strikingly noticeable in super figures of the broad massive type with Phidian lines. The addition of light drapery, however, converts the apparent faults into virtues, for the artist is enabled therewith to give new sweeping curves to the forms which conspicuously enhance the general beauty of the figure.