"Háček" (the hook, as over the "c" in the word itself ) did not exist in the past. Instead of it, another alphabetic character was used behind the letter.
Example : the surname ©anda.
You will likely have the transcription "Sanda". Possible forms in the manuscripts are Szanda, Ssanda, ©sanda, and, of course, Schanda, in the records written in German.
Some letters were not used. In the past, "g" was used instean of today's "j" ( as "j" was written on place of "i"). Have this in mind when researching indexes.
Example : researching the surname "Jelinek", you must look for "Gelinek" also.
Manytimes, the phonetic form was used. This way, the surnames "Koktavy" and "Kochtavy" were one and the same name (e.g. in 1690).
Example : my surname, Fiala :
you can see it written as "Figala", as the pronounciation is [: fijala : ] (j as in "few", not in "John"), and the "j" - that means, "g" - might have been inserted.
In the very past times, the surnames which are different now, might have been used interchangeably for one and the same family ( e.g. Holý - Holík )
Basicaly, the female form is created by adding the suffix "-ová" to the surname, omitting the last letter, provided that it is a vowel ( Just -> Justová, Krofta -> Kroftová ). Also, the "e" before the last consonant (but n,l,r ! ) is omitted : ( Turek -> Turková, Pazdera -> Pazderová)
If the surname itself has the meaning of an adjective, the suffix -ová is not used, and the female form of this adjective is used : (Hrubý - Hrubá ). If the meaning of the surname is some substantive, which declension is according to the paradigm of adjectives, the female form is created accordingly. In some cases, it remains unchanged, due to this rule (Krejčí -> Krejčí) You can see the form "Krejčová" of course... nevertheless, if you understand all written above, you already know that it was derived from "Krejča", and that the difference between "Krejča" and "Krejčí" was not too significant in the past.
Sometimes it is not possible to deduce the male form from the female one exactly. Thus, the form "Bílková" can point to any of the surnames Bílek, Bílka, Bílko, and you cannot know whether the father of Miss Jančová was Mr Janč or Mr Janča.
You can meet the possesive form of the female form also, in the old records. The surname Kuba has the correct female form Kubová, but you can find Kubů or even Kubových. These possesive forms were used for men, as well. Another similar form can be found in records written in old German language - Sladkin instead of Sladkova (from Sladek).
On the contrary to English, the surnames in Czech has different forms in different grammatical case. It's not possible to discuss it exactly in the extent of this page, but one example is worth mentioning, as it is a source of common mistake :
"Anna, dcera Jana Pazdery" means " Ann, daughter of Jan Pazdera".
Once you have discovered the place of origin of your ancestors, everything should be possible to be found. Do not bother about the spelling too much. It might be different in the past. The name can be misfigured. You are not always able to guess the nominative.
I may post here some advices in future ... but in fact, when you are in doubt, the best advice is to ask a professional. He or she can use some of the books containing the exact history of the geographical names. And, he or she should be more experienced in ... guessing right.
Example : you have the place of origin "Bejralovice". You are looking thru the net, including Jewishgen, and the only result is, that you learn that there is no such a place. Well : "-ovice" is a suffix of many places, as well as "-ov" ( both of them correspond to the ethymology). "B" and "V" ... local name ... might be used one instead of another, eventualy. And the "r" ? If you are familiar with the old manuscripts, it's no problem for you. Simply a mistake in the transcription, it's no "r", but "v"... they can look nearly the same. You have "Vejvalov" or "Vývalov" now. One more step ... and you are home. Vejvanov. It's a true case from my practise, no tale. (Surely, it was not as easy as I described here, as more variants had to be considered.)