I have a set of values in a range of cells on a worksheet, Patterns!B2:B375

There are also values in the corresponding cells in nearby columns,
such as Patterns!C2:C375, Patterns!D2:D375, Patterns!E2:E375, etc.

I want to calculate the correlations between the values in the first range,
and the values in each of the other columns. And I want these correlation
values to appear on different rows (for example, in place of the ### below):

Corrleation
ColC ###
ColD ###
ColE ###

And so, I thought I could use this function:

=CORREL(INDIRECT("Patterns!R2C"&(ROW()+1)&":R375C" &(ROW()+1),FALSE),Patterns!B$2:B$375)

in each of the cells, and it would automatically have each row calculate the
correlation with a different column of values.

But... it doesn't work. It gives me #N/A.

The TRULY STRANGE thing about this, is if I simply substitute the actual row
number in place of the ROW() function, it works:

=CORREL(INDIRECT("Patterns!R2C"&(2+1)&":R375C"&(2+ 1),FALSE),Patterns!B$2:B$375)

This actually produces the correct numerical result. And all I did, was
replace the embedded row() function call with an actual number.

Why should I not be able to embed row() within indirect()?

---G