Some more info...

We removed the Philips MF RC531 chip and started checking the signals of the ATMEL chip (during external access) at Oscilloscope level. Then we found something different.

The code for writeregister is
P4 &= ~(0x10); //Chip Select enable
AUXR |= 0x02; // Select external data
XBYTE [byAddress] = byValue;
AUXR &= ~(0x02);
P4 |= 0x10; //Chip Select disable

Code for readregister is
BYTE byRetVal = 0;
P4 &= ~(0x10); //Chip Select enable
AUXR |= 0x02; // Select external data
byRetVal = XBYTE [byAddress];
AUXR &= ~(0x02);
P4 |= 0x10; //Chip Select disable
return (byRetVal);

And what we found in the scope was
1) During a writeregister the chip select is proper and there is a WR pulse within the chip select which means a proper write pulse out from the ATMEL chip

2) During the readregister, however the chip select is proper, but there is no RD pulse within the chip select. A very meagre spike alone is seen. Why is the chip behaving differently. Is there anything special to be done for reading alone (Any config registers or anything)

Thanks for the support, looking forward for more..