I explained this to you once before.
It takes 32 bits to encode a floating point number. %R registers in the 90-30 are 16 bits. When you program a function that uses REAL data type Logicmaster takes two consecutive %R addresses to encode the REAL number. In your case, it uses %R00010 and %R00011.
The reference table display shows the contents of each individual sixteen bit %R register. You can make the display in the reference table match the display on the program screen. With the cursor in the position shown in "R10.jpg", press F4. Logicmaster will then interpret the data in both %R10 and %R11 as a floating pont number and display it that way.