Nor need we wonder that this portion of their creed was not divulged by the Persians to Herodotus or to Xenophon, since it is exactly the sort of subject on which reticence was natural and might have been anticipated.

Neither the lively Halicarnassian, nor the pleasant but somewhat shallow Athenian, had the gift of penetrating very deeply into the inner mind of a foreign people; added to which, it is to be remembered that they were unacquainted with Persia Proper, and drew their knowledge of Persian opinions and customs either from hearsay or from the creed and practices of the probably mixed garrisons which held Asia Minor, Syria, and Egypt.
Persian worship, in these early times, was doubtless that enjoined by the Zendavesta, comprising prayer and thanksgiving to Ormazd and the good spirits of his creation, the recitation of Gathas or hymns, the performance of sacrifice, and participation in the Soma ceremony.
Worship seems to have taken place in temples, which are mentioned (according to the belief of most cuneiform scholars) in the Behistun inscription.

Of the character of these buildings we can say nothing.
It has been thought that those two massive square towers so similar in construction, which exist in a more or less ruined condition at Murgab and Nakhsh-i-Rustam, are Persian temples of the early period, built to contain an altar on which the priests offered victims.

But the absence of any trace of an altar from both, the total want of religious emblems, and the extremely small size of the single apartment which each tower contains, make strongly against the temple theory; not to mention that a much more probable use may be suggested for the buildings.
With respect to the altars upon which sacrifice was offered, we are not left wholly without evidence.

The Persian monarchs of the early period, including Darius Hystaspis, represented themselves on their tombs in the act of worship.