I have googled 225.225.225.0 used as a subnet mask.
There does seem to be a lot of stuff, but in many cases people had just entered it wrongly (probably a bit of dyslexia).
I screen-grabbed such an occasion from a help forum - one of many similar occurrances on the net.
I still maintain that using 225's is such an odd mask pattern that it has to be a case of number-blindness - you have even responded wrongly in your last post - 255 is the norm, 225 is the odd-ball.
Refer to my previous post where I mapped out the numbers in binary. It is customary (but not mandatory) to use a series of contiguous "1"s from left to right to define the NetID part of the address.
examples
255 - 11111111
254 - 11111110
252 - 11111100
248 - 11111000
240 - 11110000
224 - 11100000
225 - 11100001
192 - 11000000
128 - 10000000
000 - 00000000
Doing so enables us mere humans to do the maths easier to determine if two IPs are on the same subnet.
The hardware is exceptionally good at it, as it simply AND's the two IP addresses with the subnet mask, and if the two results are the same, then they are on the same subnet, and the hosts with those IP addresses are allowed to communicate with each other.
IMHO - although technically using 225.225.225.0 is indeed valid (the hardware won't care about the odd-ball bit patterns), it does produce such weird combinations of allowable IP addresses on the subnet that it has to be a "number error"